These changes add the ability to export a stack's latest deployment or
import a new deployment to a stack via the Pulumi CLI. These
capabilities are exposed by two new verbs under `stack`:
- export, which writes the current stack's latest deployment to stdout
- import, which reads a new deployment from stdin and applies it to the
current stack.
In the local case, this simply involves reading/writing the stack's
latest checkpoint file. In the cloud case, this involves hitting two new
endpoints on the service to perform the export or import.
This PR updates the `pkg/testing/integration` package to support running integration tests against the Pulumi Service if desired. This is done through adding new options to `ProgramTestOptions`. (Generally adding support for providing values to flags that were previously inaccessible.)
I added an integration test to confirm that it all works if the PULUMI_API environment variable is set. These tests aren't run in Travis, only manually. Since we cannot reliably run tests from `master` against the service because of the delay in rolling out updates to the Pulumi SDK, etc.
This change incorporates feedback on https://github.com/pulumi/pulumi/pull/764,
in addition to refactoring the retry logic to use our retry framework rather
than hand-rolling it in the REST API code. It's a minor improvement, but at
least lets us consolidate some of this logic which we'll undoubtedly use more
of over time.
We saw an issue where a user was mid-update, and got a networking
error stating `read: operation timed out`. We believe this was simply
a local client error, due to a flaky network. We should be resilient
to such things during updates, particularly when there's no way to
"reattach" to an in-progress udpate (see pulumi/pulumi#762).
This change accomplishes this by changing our retry logic in the
cloud backend's waitForUpdates function. Namely:
* We recognize three types of failure, and react differently:
- Expected HTTP errors. For instance, the 504 Gateway Timeouts
that we already retried in the face of. In these cases, we will
silently retry up to 10 times. After 10 times, we begin warning
the user just in case this is a persistent condition.
- Unexpected HTTP errors. The CLI will quit immediately and issue
an error to the user, in the usual ways. This covers
Unauthorized among other things. Over time, we may find that we
want to intentionally move some HTTP errors into the above.
- Anything else. This covers the transient networking errors case
that we have just seen. I'll admit, it's a wide net, but any
instance of this error issues a warning and it's up to the user
to ^C out of it. We also log the error so that we'll see it if
the user shares their logs with us.
* We implement backoff logic so that we retry very quickly (100ms)
on the first failure, and more slowly thereafter (1.5x, up to a max
of 5 seconds). This helps to avoid accidentally DoSing our service.
This merging causes similar issues to those it did in `Check`, and
differs from the approach we take to `Diff`. This can causes problems
such as an inability to remove properties.
Use the new {en,de}crypt endpoints in the Pulumi.com API to secure
secret config values. The ciphertext for a secret config value is bound
to the stack to which it applies and cannot be shared with other stacks
(e.g. by copy/pasting it around in Pulumi.yaml). All secrets will need
to be encrypted once per target stack.
This fixes a few more edit directory issues, where we didn't
correctly propagate the changes in edit directory required during
subsequent destroy/stack activities. It also fixes a few error
paths so that we preserve the right directory to be removed.
This change implements resource protection, as per pulumi/pulumi#689.
The overall idea is that a resource can be marked as "protect: true",
which will prevent deletion of that resource for any reason whatsoever
(straight deletion, replacement, etc). This is expressed in the
program. To "unprotect" a resource, one must perform an update setting
"protect: false", and then afterwards, they can delete the resource.
For example:
let res = new MyResource("precious", { .. }, { protect: true });
Afterwards, the resource will display in the CLI with a lock icon, and
any attempts to remove it will fail in the usual ways (in planning or,
worst case, during an actual update).
This was done by adding a new ResourceOptions bag parameter to the
base Resource types. This is unfortunately a breaking change, but now
is the right time to take this one. We had been adding new settings
one by one -- like parent and dependsOn -- and this new approach will
set us up to add any number of additional settings down the road,
without needing to worry about breaking anything ever again.
This is related to protected stacks, as described in
pulumi/pulumi-service#399. Most likely this will serve as a foundational
building block that enables the coarser grained policy management.
This change adds the ability to do very coarse-grained negative
tests in our integration test framework. Either a test itself,
or one of its edits, may be marked ExpectFailure == true, at which
point either the preview or update MUST fail (and, if one fails
without this being set, we still treat it as an error).
Our recent changes to colorization changed from a boolean to a tri-valued
enum (Always, Never, Raw). The events from the service, however, are still
boolean-valued. This changes the message payload to carry the full values.
The prior behavior with cloud authentication was a bit confusing
when authenticating against anything but https://pulumi.com/. This
change fixes a few aspects of this:
* Improve error messages to differentiate between "authentication
failed" and "you haven't logged into the target cloud URL."
* Default to the cloud you're currently authenticated with, rather
than unconditionally selecting https://pulumi.com/. This ensures
$ pulumi login -c https://api.moolumi.io
$ pulumi stack ls
works, versus what was currently required
$ pulumi login -c https://api.moolumi.io
$ pulumi stack ls -c https://api.moolumi.io
with confusing error messages if you forgot the second -c.
* To do this, our default cloud logic changes to
1) Prefer the explicit -c if supplied;
2) Otherwise, pick the "currently authenticated" cloud; this is
the last cloud to have been targeted with pulumi login, or
otherwise the single cloud in the list if there is only one;
3) https://pulumi.com/ otherwise.
We hadn't previously passed the planning flag when printing resource
outputs, meaning any computed ones now are being printed as "undefined".
Instead, we prefer to see the "computed<string>" type name.
We do not need all of the information in the old state for this call, as
outputs will not be read by the provider during validation or defaults
computation.
This change adds a bit more tracing context to RPC marshaling
logging so that it's easier to attribute certain marshaling calls.
Prior to this, we'd just have a flat list of "marshaled property X"
without any information about what the marshaling pertained to.
This change passes a resource's old output state, so that it contains
everything -- defaults included -- for purposes of the provider's diffing.
Not doing so can lead the provider into thinking some of the requisite
state is missing.
Part of the work to make it easier to tests of diff output. Specifically, we now allow users to pass --color=option for several pulumi commands. 'option' can be one of 'always', 'never', 'raw', and 'auto' (the default).
The meaning of these flags are:
1. auto: colorize normally, unless in --debug
2. always: always colorize no matter what
3. never: never colorize no matter what.
4. raw: colorize, but preserve the original "<{%%}>" style control codes and not the translated platform specific codes. This is for testing purposes and ensures we can have test for this stuff across platform.
This will allow us to remove a lot of current boilerplate in individual tests, and move it into the test harness.
Note that this will require updating users of the integration test framework. By moving to a property bag of inputs, we should avoid needing future breaking changes to this API though.
This change adds rudimentary delete-before-create support (see
pulumi/pulumi#450). This cannot possibly be complete until we also
implement pulumi/pulumi#624, becuase we may try to delete a resource
while it still has dependent resources (which almost certainly will
fail). But until then, we can use this to manually unwedge ourselves
for leaf-node resources that do not support old and new resources
living side-by-side.
As articulated in #714, the way config defaults to workspace-local
configuration is a bit error prone, especially now with the cloud
workflow being the default. This change implements several improvements:
* First, --save defaults to true, so that configuration changes will
persist into your project file. If you want the old local workspace
behavior, you can specify --save=false.
* Second, the order in which we applied configuration was a little
strange, because workspace settings overwrote project settings.
The order is changed now so that we take most specific over least
specific configuration. Per-stack is considered more specific
than global and project settings are considered more specific
than workspace.
* We now warn anytime workspace local configuration is used. This
is a developer scenario and can have subtle effects. It is simply
not safe to use in a team environment. In fact, I lost an arm
this morning due to workspace config... and that's why you always
issue warnings for unsafe things.
If the service returns a 504, we happily keep looping around and
retrying until we get a valid update. Unfortunately, we missed the
else condition, which is what happens when this isn't a 504, leading
us to swallow real errors (500 and the like). Trivial fix.
Fixespulumi/pulumi#712.
Rather than displaying `computed` as we do during previews, display
`undefined` instead, as these values may be intentionally undefined by
the user's program.
Fixes#633.
Previously, we would compute all the output from the preview operation
and then display it as a single event. Instead, we should do this how
we handle things during deploy. Each logical print operaton gets a
single event and we stream them back over a channel during planning.
Fixes#660
This change just flows the project's "main" directory all the way
through to the plugins, fixing #667. In that work item, we discussed
alternative approaches, such as rewriting the asset paths, but this
is tricky because it's very tough to do without those absolute paths
somehow ending up in the checkpoint files. Just launching the
processes with the right pwd is far easier and safer, and it turns
out that, conveniently, we set up the plugin context in exactly the
same place that we read the project information.
If the stack config specifies AWS credentials, these should be used in the operations provider instead of ambient credentials. This is necessary to ensure that we have access to resources in the target account.
Fixespulumi/pulumi-service#389
* Take an options pointer so values can change as a test runs.
* Don't pass redundant information.
* Extract initialization routine.
* Fix caller.
* Check return value.
* Extract destruction logic.
* Move preview and update into their own function.
* Inline null check.
This change adds a `pulumi stack output` command. When passed no
arguments, it prints all stack output properties, in exactly the
same format as `pulumi stack` does (just without all the other stuff).
More importantly, if you pass a specific output property, a la
`pulumi stack output clusterARN`, just that property will be printed,
in a scriptable-friendly manner. This will help us automate wiring
multiple layers of stacks together during deployments.
This fixespulumi/pulumi#659.
* Revert "Make sure we properly update dir so that pulumi-destroy works."
This reverts commit 56bfc57998.
* Revert "Edits needs to continuously pass along the new directory. (#668)"
This reverts commit 8bd1822722.
* Revert "Refactor test code to make it simpler to validate code in the middle. (#662)"
This reverts commit ed65360157.
If a cloud you've previously authenticated with goes away -- as ours
sort of did, because the cloud endpoing in the CLI changed (to actually
be correct) -- then you can't logout without manually editing the
credentials file in your workspace. This is a little annoying. So,
rather than that, let's have a `pulumi logout --all` command that just
logs out of all clouds you are presently authenticated with.
This changes two things:
1) Always show the root stack, even when unchanged. This ensures
that you see the outputs during any updates, etc., which is nice.
2) Always show all outputs after operations. Before, we did some
diffing between old/new, which actually doesn't make much sense.
- Fix#644 "Re-enable use of local dev stacks". Rather than trying to be smart about using "pulumi.com" and inferring "api.pulumi.com", we just use whatever cloud URL the user provides. (e.g. "http://localhost:8080") We can improve the user experience later by providing friendly names for these things upon login. So we can show "Pulumi" or "Contoso" instead of the URL -- but let's cross that bridge later.
- Fix#640, "Better error on `pulumi login`". We only provide the more friendly error about needing to login IFF you are trying to use an authenticated Pulumi API without having creds.
The two-phase output properties change broke the ability to recover
from a failed replacement that yields pending deletes in the checkpoint.
The issue here is simply that we should remember pending registrations
only for logical operations that *also* have a "new" state (create or
update). This commit fixes this, and also adds a new step test with
fault injection to probe many interesting combinations of steps.
If the process we are trying to kill has already exited, don't treat
this as an error. This can happen when we snapshot the process tree
before the process exits but it has exited by the time we get to
trying to kill it.
Fixes#654
Every single resource has a type prefix of
pulumi:pulumi:Stack$
which makes URNs quite lengthy without adding any value. Since
they all have this prefix, adding it doesn't help to disambiguate.
This change skips adding the parent URN part when it is the built-in
automatic stack type name.
At some point, we fixed a bug in the way state is managed for "same"
steps, which meant that we wouldn't see newly added output properties.
This had the effect that, if you had a stack already stood up, and
updated it to have output properties, we would miss them. (Stacks
stood up from scratch would still have them.) This fixes that problem,
in addition to two other things: 1) we need to sort output property
names to ensure a deterministic ordering, and 2) we need to also
unconditionally apply the outputs RPC coming in, to ensure that the
resulting resource always has the correct outputs (so that for example
deleting prior output properties actually deletes them).
Also add some testing for this area to make sure we don't break again.
Fixespulumi/pulumi#631.
These changes push the `config.{Map,Value}` interfaces further down into
the deployment engine so that configuration can be decrypted nearer to
its use.
This is the first part of the fix for pulumi/pulumi-ppc#112.
As documented in issue #616, the inputs/defaults/outputs model we have
today has fundamental problems. The crux of the issue is that our
current design requires that defaults present in the old state of a
resource are applied to the new inputs for that resource.
Unfortunately, it is not possible for the engine to decide which
defaults remain applicable and which do not; only the provider has that
knowledge.
These changes take a more tactical approach to resolving this issue than
that originally proposed in #616 that avoids breaking compatibility with
existing checkpoints. Rather than treating the Pulumi inputs as the
provider input properties for a resource, these inputs are first
translated by `Check`. In order to accommodate provider defaults that
were chosen for the old resource but should not change for the new,
`Check` now takes the old provider inputs as well as the new Pulumi
inputs. Rather than the Pulumi inputs and provider defaults, the
provider inputs returned by `Check` are recorded in the checkpoint file.
Put simply, these changes remove defaults as a first-class concept
(except inasmuch as is required to retain the ability to read old
checkpoint files) and move the responsibilty for manging and
merging defaults into the provider that supplies them.
Fixes#616.
This change implements some feedback from @ellismg.
* Make backend.Stack an interface and let backends implement it,
enabling dynamic type testing/casting to access information
specific to that backend. For instance, the cloud.Stack conveys
the cloud URL, org name, and PPC name, for each stack.
* Similarly expose specialized backend.Backend interfaces,
local.Backend and cloud.Backend, to convey specific information.
* Redo a bunch of the commands in terms of these.
* Keeping with this theme, turn the CreateStack options into an
opaque interface{}, and let the specific backends expose their
own structures with their own settings (like PPC name in cloud).
* Show both the org and PPC names in the cloud column printed in
the stack ls command, in addition to the Pulumi Cloud URL.
Unrelated, but useful:
* Special case the 401 HTTP response and make a friendly error,
to tell the developer they must use `pulumi login`. This is
better than tossing raw "401: Unauthorized" errors in their face.
* Change the "Updating stack '..' in the Pulumi Cloud" message to
use the correct action verb ("Previewing", "Destroying", etc).
This change does two things:
1) Adds progress reporting to our uploads.
2) Eliminate the sleeps that burned 7 seconds at the front of
any cloud update, needlessly. It's actually impressively
fast without these!
This improves the overall cloud CLI experience workflow.
Now whether a stack is local or cloud is inherent to the stack
itself. If you interact with a cloud stack, we transparently talk
to the cloud; if you interact with a local stack, we just do the
right thing, and perform all operations locally. Aside from sometimes
seeing a cloud emoji pop-up ☁️, the experience is quite similar.
For example, to initialize a new cloud stack, simply:
$ pulumi login
Logging into Pulumi Cloud: https://pulumi.com/
Enter Pulumi access token: <enter your token>
$ pulumi stack init my-cloud-stack
Note that you may log into a specific cloud if you'd like. For
now, this is just for our own testing purposes, but someday when we
support custom clouds (e.g., Enterprise), you can just say:
$ pulumi login --cloud-url https://corp.acme.my-ppc.net:9873
The cloud is now the default. If you instead prefer a "fire and
forget" style of stack, you can skip the login and pass `--local`:
$ pulumi stack init my-faf-stack --local
If you are logged in and run `pulumi`, we tell you as much:
$ pulumi
Usage:
pulumi [command]
// as before...
Currently logged into the Pulumi Cloud ☁️https://pulumi.com/
And if you list your stacks, we tell you which one is local or not:
$ pulumi stack ls
NAME LAST UPDATE RESOURCE COUNT CLOUD URL
my-cloud-stack 2017-12-01 ... 3 https://pulumi.com/
my-faf-stack n/a 0 n/a
And `pulumi stack` by itself prints information like your cloud org,
PPC name, and so on, in addition to the usuals.
I shall write up more details and make sure to document these changes.
This change also fairly significantly refactors the layout of cloud
versus local logic, so that the cmd/ package is resonsible for CLI
things, and the new pkg/backend/ package is responsible for the
backends. The following is the overall resulting package architecture:
* The backend.Backend interface can be implemented to substitute
a new backend. This has operations to get and list stacks,
perform updates, and so on.
* The backend.Stack struct is a wrapper around a stack that has
or is being manipulated by a Backend. It resembles our existing
Stack notions in the engine, but carries additional metadata
about its source. Notably, it offers functions that allow
operations like updating and deleting on the Backend from which
it came.
* There is very little else in the pkg/backend/ package.
* A new package, pkg/backend/local/, encapsulates all local state
management for "fire and forget" scenarios. It simply implements
the above logic and contains anything specific to the local
experience.
* A peer package, pkg/backend/cloud/, encapsulates all logic
required for the cloud experience. This includes its subpackage
apitype/ which contains JSON schema descriptions required for
REST calls against the cloud backend. It also contains handy
functions to list which clouds we have authenticated with.
* A subpackage here, pkg/backend/state/, is not a provider at all.
Instead, it contains all of the state management functions that
are currently shared between local and cloud backends. This
includes configuration logic -- including encryption -- as well
as logic pertaining to which stacks are known to the workspace.
This addresses pulumi/pulumi#629 and pulumi/pulumi#494.
This change adds a new manifest section to the checkpoint files.
The existing time moves into it, and we add to it the version of
the Pulumi CLI that created it, along with the names, types, and
versions of all plugins used to generate the file. There is a
magic cookie that we also use during verification.
This is to help keep us sane when debugging problems "in the wild,"
and I'm sure we will add more to it over time (checksum, etc).
For example, after an up, you can now see this in `pulumi stack`:
```
Current stack is demo:
Last updated at 2017-12-01 13:48:49.815740523 -0800 PST
Pulumi version v0.8.3-79-g1ab99ad
Plugin pulumi-provider-aws [resource] version v0.8.3-22-g4363e77
Plugin pulumi-langhost-nodejs [language] version v0.8.3-79-g77bb6b6
Checkpoint file is /Users/joeduffy/dev/code/src/github.com/pulumi/pulumi-aws/.pulumi/stacks/webserver/demo.json
```
This addresses pulumi/pulumi#628.
This change introduces automatic integrity checking for snapshots.
Hopefully this will help us track down what's going on in
pulumi/pulumi#613. Eventually we probably want to make this opt-in,
or disable it entirely other than for internal Pulumi debugging, but
until we add more complete DAG verification, it's relatively cheap
and is worthwhile to leave on for now.
This PR just wires the `Package.Main` field to the Pulumi Service (and in subsequent PRs, the `pulumi-service` and `pulumi-ppc` repos).
@joeduffy , should we just upload the entire `package.Package` type with the `UpdateProgramRequest` type? I'm not sure we want to treat that type as part of part of our public API surface area. But on the other hand, we'll need to mirror relevant fields in N places if we don't.
This change adds prefix +/- change markers for added and deleted
lines, in addition to the existing colorization. It also elides
empty strings from the diff which, due to our newline splitting,
always ended up with extra whitespace in the +/- output.
Previously, we were inconsistent on how we handled argument validation
in the CLI. Many commands used cobra.Command's Args property to
provide a validator if they took arguments, but commands which did not
rarely used cobra.NoArgs to indicate this.
This change does two things:
1. Introduce `cmdutil.ArgsFunc` which works like `cmdutil.RunFunc`, it
wraps an existing cobra type and lets us control the behavior when an
arguments validator fails.
2. Ensure every command sets the Args property with an instance of
cmdutil.ArgsFunc. The cmdutil package defines wrapers for all the
cobra validators we are using, to prevent us from having to spell out
`cmduitl.ArgsFunc(...)` everywhere.
Fixes#588
The prior change was incorrectly handling snapshotting of replacement
operations. Further, in hindsight, the older model of having steps
manage their interaction with the snapshot marking was clearer, so
I've essentially brought that back, merging it with the other changes.
This change simplifies the necessary RPC changes for components.
Instead of a Begin/End pair, which complicates the whole system
because now we have the opportunity of a missing End call, we will
simply let RPCs come in that append outputs to existing states.
We need to invoke the post-step event hook *after* updating the
state snapshots, so that it will write out the updated state.
We also need to re-serialize the snapshot again after we receive
updated output properties, otherwise they could be missing if this
happens to be the last resource (e.g., as in Stacks).
This change brings back component outputs to the overall system again.
In doing so, it generally overhauls the way we do resource RPCs a bit:
* Instead of RegisterResource and CompleteResource, we call these
BeginRegisterResource and EndRegisterResource, which begins to model
these as effectively "asynchronous" resource requests. This should also
help with parallelism (https://github.com/pulumi/pulumi/issues/106).
* Flip the CLI/engine a little on its head. Rather than it driving the
planning and deployment process, we move more to a model where it
simply observes it. This is done by implementing an event handler
interface with three events: OnResourceStepPre, OnResourceStepPost,
and OnResourceComplete. The first two are invoked immediately before
and after any step operation, and the latter is invoked whenever a
EndRegisterResource comes in. The reason for the asymmetry here is
that the checkpointing logic in the deployment engine is largely
untouched (intentionally, as this is a sensitive part of the system),
and so the "begin"/"end" nature doesn't flow through faithfully.
* Also make the engine more event-oriented in its terminology and the
way it handles the incoming BeginRegisterResource and
EndRegisterResource events from the language host. This is the first
step down a long road of incrementally refactoring the engine to work
this way, a necessary prerequisite for parallelism.
This changes our resource creation logic to tolerate missing children.
This ultimately shouldn't be necessary, but we appear to have a bug where
sometimes during deletions we end up with dangling child pointers. See
pulumi/pulumi#613; it tracks fixing the bug, and also removing this
workaround and reenabling the assertion.
When merging inputs and defaults in order to construct the set of inputs
for a call to `Create`, we must recursively merge each property value:
the provided defaults may contain nested values that must be present in
the merged result.
This change includes a NewResourceMap API, alongside the existing,
but renamed, NewResourceTree API. This also tidies up the resulting
struct for public consumption. This is required for downstream tests.
* Don't show +s, -s, and ~s deeply. The intended format here looks
more like
+ aws:iam/instanceProfile:InstanceProfile (create)
[urn=urn:pulumi:test::aws/minimal::aws/iam/instanceProfile:InstanceProfile::ip2]
name: "ip2-079a29f428dc9987"
path: "/"
role: "ir-d0a632e3084a0252"
versus
+ aws:iam/instanceProfile:InstanceProfile (create)
+ [urn=urn:pulumi:test::aws/minimal::aws/iam/instanceProfile:InstanceProfile::ip2]
+ name: "ip2-079a29f428dc9987"
+ path: "/"
+ role: "ir-d0a632e3084a0252"
This makes it easier to see the resources modified in the output.
* Print adds/deletes during updates as
- property: "x"
+ property: "y"
rather than
~ property: "x"
~ property: "y"
the latter of which doesn't really tell you what's new/old.
* Show parent indentation on output properties, so they line up correctly.
* Only print stack outputs if not undefined.
This change switches from child lists to parent pointers, in the
way resource ancestries are represented. This cleans up a fair bit
of the old parenting logic, including all notion of ambient parent
scopes (and will notably address pulumi/pulumi#435).
This lets us show a more parent/child display in the output when
doing planning and updating. For instance, here is an update of
a lambda's text, which is logically part of a cloud timer:
* cloud:timer:Timer: (same)
[urn=urn:pulumi:malta::lm-cloud:☁️timer:Timer::lm-cts-malta-job-CleanSnapshots]
* cloud:function:Function: (same)
[urn=urn:pulumi:malta::lm-cloud:☁️function:Function::lm-cts-malta-job-CleanSnapshots]
* aws:serverless:Function: (same)
[urn=urn:pulumi:malta::lm-cloud::aws:serverless:Function::lm-cts-malta-job-CleanSnapshots]
~ aws:lambda/function:Function: (modify)
[id=lm-cts-malta-job-CleanSnapshots-fee4f3bf41280741]
[urn=urn:pulumi:malta::lm-cloud::aws:lambda/function:Function::lm-cts-malta-job-CleanSnapshots]
- code : archive(assets:2092f44) {
// etc etc etc
Note that we still get walls of text, but this will be actually
quite nice when combined with pulumi/pulumi#454.
I've also suppressed printing properties that didn't change during
updates when --detailed was not passed, and also suppressed empty
strings and zero-length arrays (since TF uses these as defaults in
many places and it just makes creation and deletion quite verbose).
Note that this is a far cry from everything we can possibly do
here as part of pulumi/pulumi#340 (and even pulumi/pulumi#417).
But it's a good start towards taming some of our output spew.
The first exception relates to how we launch plugins. Plugin paths are
calculated using a well-known set of rules; this makes `gas` suspicious
due to the need to use a variable to store the path of the plugin.
The second and third are in test code and aren't terribly concerning.
The latter exception asks `gas` to ignore the access key we hard-code
into the integration tests for our Pulumi test account.
The fourth exception allows use to use more permissive permissions for
the `.pulumi` directory than `gas` would prefer. We use `755`; `gas`
wants `700` or stricter. `755` is the default for `mkdir` and `.git` and
so seems like a reasonable choice for us.
By default, debugging events are not displayed by `pulumi`; the `-d`
flag must be provided to enable their display. This is necessary e.g.
to view debugging output from Terraform when TF logging is enabled.
These changes add the option to display debug output from `pulumi` to
the integration test framework.
These changes also contain a small fix for the display of component
children.
Parallelize collection of logs at each layer of the resource tree walk. Since almost all cost of log collection is at leaf nodes, this should allow the total time for log collection to be close to the time taken for the single longest leaf node, instead of the sum of all leaf nodes.
We need to clear the resource filter during the resource tree walk to ensure that logs from children of matched resources are collected and aggregated.
We have to reverse engineer the name from the soruce LogGroup information since that is all we got at runtime, but luckily that is sufficient given current name generation approach.
This kind of code is *very* sensitive to any changes to automatic name generation - but that is likely inevitable at this layer.
Outside of `.pulumiignore` we support a few "default" excludes that
try to push folks towards a pit of succes.
Previously, there was no way to opt out of these, which would be bad
if our huristics caused something youto really care about to be
elided. With this change, we add an optional setting in Pulumi.yaml
that allows you to opt out of this behavior.
As part of the work, I changed .git to be one of these "default"
excludes instead of it only happening if you had a .pulumiignore file
in a directory
Previously, we would archive every file in node_modules, however we
only actually needed the production modules. This change adds an
implicit ignorer when walking a directory that has a package.json file
in to exclude stuff under `node_modules` which is not pulled in via an
entry in the `dependencies` section of the package.json.
This change will also cause the archives that we upload to not include
either pulumi or any @pulumi/... packages, since they are not listed
in the dependencies section ofa package.json. This is fine, since we
linked in the versions we have in the cloud anyway. When we move to
using npm for these packages (instead of linking) then they will be
included.
This won't be needed once pulumi/pulumi-ppc#95 has landed and we've
updated our PPCs, but for now always add directory entires (even if
the files would be excluded) for every directory.
When deploying a project via the Pulumi.com service, we have to upload
the entire "context" of your project to Pulumi.com. The context of the
program is all files in the directory tree rooted by the `Pulumi.yaml`
file, which will often contain stuff we don't want to upload, but
previously we had no control over what would be updated (and so folks
would do hacky things like delete folders before running `pulumi
update`).
This change adds support for `.pulumiignore` files which should behave
like `.gitignore`. In addition, we were not previously compressing
files when we added them to the zip archive we uploaded and now.
By default, every .pulumiignore file is treated as if it had an
exclusion for `.git/` at the top of the file (users can override this
by adding an explicit `!.git/` to their file) since it is very
unlikely for there to ever be a reason to upload the .git folder to
the service.
Fixespulumi/pulumi-service#122
This change adds back component output properties. Doing so
requires splitting the RPC interface for creating resources in
half, with an initial RegisterResource which contains all of the
input properties, and a final CompleteResource which optionally
contains any output properties synthesized by the component.
This change switches from child lists to parent pointers, in the
way resource ancestries are represented. This cleans up a fair bit
of the old parenting logic, including all notion of ambient parent
scopes (and will notably address pulumi/pulumi#435).
This lets us show a more parent/child display in the output when
doing planning and updating. For instance, here is an update of
a lambda's text, which is logically part of a cloud timer:
* cloud:timer:Timer: (same)
[urn=urn:pulumi:malta::lm-cloud:☁️timer:Timer::lm-cts-malta-job-CleanSnapshots]
* cloud:function:Function: (same)
[urn=urn:pulumi:malta::lm-cloud:☁️function:Function::lm-cts-malta-job-CleanSnapshots]
* aws:serverless:Function: (same)
[urn=urn:pulumi:malta::lm-cloud::aws:serverless:Function::lm-cts-malta-job-CleanSnapshots]
~ aws:lambda/function:Function: (modify)
[id=lm-cts-malta-job-CleanSnapshots-fee4f3bf41280741]
[urn=urn:pulumi:malta::lm-cloud::aws:lambda/function:Function::lm-cts-malta-job-CleanSnapshots]
- code : archive(assets:2092f44) {
// etc etc etc
Note that we still get walls of text, but this will be actually
quite nice when combined with pulumi/pulumi#454.
I've also suppressed printing properties that didn't change during
updates when --detailed was not passed, and also suppressed empty
strings and zero-length arrays (since TF uses these as defaults in
many places and it just makes creation and deletion quite verbose).
Note that this is a far cry from everything we can possibly do
here as part of pulumi/pulumi#340 (and even pulumi/pulumi#417).
But it's a good start towards taming some of our output spew.
Adds support for top-level exports in the main script of a Pulumi Program to be captured as stack-level output properties.
This create a new `pulumi:pulumi:Stack` component as the root of the resource tree in all Pulumi programs. That resources has properties for each top-level export in the Node.js script.
Running `pulumi stack` will display the current value of these outputs.
A handful of UX improvments for config:
- `pulumi config ls` has been removed. Now, `pulumi config` with no
arguments prints the table of configuration values for a stack and
a new command `pulumi config get <key>` prints the value for a
single configuration key (useful for scripting).
- `pulumi config text` and `pulumi config secret` have been merged
into a single command `pulumi config set`. The flag `--secret` can
be used to encrypt the value we store (like `pulumi config secret`
used to do).
- To make it obvious that setting a value with `pulumi config set` is
in plan text, we now echo a message back to the user saying we
added the configuration value in plaintext.
Fixes#552
This change fixes getProject to return the project name, as
originally intended. (One line was missing.)
It also adds an integration test for this.
Fixespulumi/pulumi#580.
Because the Pulumi.yaml file demarcates the boundary used when
uploading a program to the Pulumi.com service at the moment, we
have trouble when a Pulumi program uses "up and over" references.
For instance, our customer wants to build a Dockerfile located
in some relative path, such as `../../elsewhere/`.
To support this, we will allow the Pulumi.yaml file to live
somewhere other than the main Pulumi entrypoint. For example,
it can live at the root of the repo, while the Pulumi program
lives in, say, `infra/`:
Pulumi.yaml:
name: as-before
main: infra/
This fixespulumi/pulumi#575. Further work can be done here to
provide even more flexibility; see pulumi/pulumi#574.
When producing a snapshot for a plan, we have two resource DAGs. One of
these is the base DAG for the plan; the other is the current DAG for the
plan. Any resource r may be present in both DAGs. In order to produce a
snapshot, we need to merge these DAGs such that all resource
dependencies are correctly preserved. Conceptually, the merge proceeds
as follows:
- Begin with an empty merged DAG.
- For each resource r in the current DAG, insert r and its outgoing
edges into the merged DAG.
- For each resource r in the base DAG:
- If r is in the merged DAG, we are done: if the resource is in the
merged DAG, it must have been in the current DAG, which accurately
captures its current dependencies.
- If r is not in the merged DAG, insert it and its outgoing edges
into the merged DAG.
Physically, however, each DAG is represented as list of resources
without explicit dependency edges. In place of edges, it is assumed that
the list represents a valid topological sort of its source DAG. Thus,
any resource r at index i in a list L must be assumed to be dependent on
all resources in L with index j s.t. j < i. Due to this representation,
we implement the algorithm above as follows to produce a merged list
that represents a valid topological sort of the merged DAG:
- Begin with an empty merged list.
- For each resource r in the current list, append r to the merged list.
r must be in a correct location in the merged list, as its position
relative to its assumed dependencies has not changed.
- For each resource r in the base list:
- If r is in the merged list, we are done by the logic given in the
original algorithm.
- If r is not in the merged list, append r to the merged list. r
must be in a correct location in the merged list:
- If any of r's dependencies were in the current list, they must
already be in the merged list and their relative order w.r.t.
r has not changed.
- If any of r's dependencies were not in the current list, they
must already be in the merged list, as they would have been
appended to the list before r.
Prior to these changes, we had been performing these operations in
reverse order: we would start by appending any resources in the old list
that were not in the new list, then append the whole of the new list.
This caused out-of-order resources when a program that produced pending
deletions failed to run to completion.
Fixes#572.
In an effort to improve performance and overall reliability, this PR moves the responsibility of uploading the Pulumi program from the Pulumi Service to the CLI. (Part of fixing https://github.com/pulumi/pulumi-service/issues/313.)
Previously the CLI would send (the dozens of MiB) program archive to the Service, which would then upload the data to S3. Now the CLI sends the data to S3 directly, avoiding the unnecessary copying of data around.
The Service-side API changes are in https://github.com/pulumi/pulumi-service/pull/323. I tested previews, updates, and destroys running the service and PPC on localhost.
The PR refactors how we handle the three kinds of program updates, and just unifies them into a single method. This makes the diff look crazy, but the code should be much simpler. I'm not sure what to do about supporting all the engine options for the Cloud-variants of Pulumi commands; I suspect that's something that should be handled at a later time.
This is the smallest possible thing that could work for both the local
development case and the case where we are targeting the Pulumi
Service.
Pulling down the pulumiframework and component packages here is a bit
of a feel bad, but we plan to rework the model soon enough to a
provider model which will remove the need for us to hold on to this
code (and will bring back the factoring where the CLI does not have
baked in knowledge of the Pulumi Cloud Framework).
Fixes#527
These changes introduce a new field, `Raw`, to `diag.Message`. This
field indicates that the contents of the message are not a format string
and should not be rendered via `Sprintf` during stringification.
The plugin std{out,err} readers have been updated to use raw messages,
and the event reader in `pulumi` has been fixed s.t. it does not format
event payloads before display.
Fixes#551.
Add the ability to upload data and timing for test runs to S3. Uploaded data is designed to be queried via a service like AWS Athena and these queries can then be imported into BI tools (Excel, QuickSight, PowerBI, etc.)
Initially hook this up to the `minimal` test as a baseline.
Note: for the purposes of this discussion, archives will be treated as
assets, as their differences are not particularly meaningful.
Currently, the identity of an asset is derived from the hash and the
location of its contents (i.e. two assets are equal iff their contents
have the same hash and the same path/URI/inline value). This means that
changing the source of an asset will cause the engine to detect a
difference in the asset even if the source's contents are identical. At
best, this leads to inefficiencies such as unnecessary updates. This
commit changes asset identity so that it is derived solely from an
asset's hash. The source of an asset's contents is no longer part of
the asset's identity, and need only be provided if the contents
themselves may need to be available (e.g. if a hash does not yet exist
for the asset or if the asset's contents might be needed for an update).
This commit also changes the way old assets are exposed to providers.
Currently, an old asset is exposed as both its hash and its contents.
This allows providers to take a dependency on the contents of an old
asset being available, even though this is not an invariant of the
system. These changes remove the contents of old assets from their
serialized form when they are passed to providers, eliminating the
ability of a provider to take such a dependency. In combination with the
changes to asset identity, this allows a provider to detect changes to
an asset simply by comparing its old and new hashes.
This is half of the fix for [pulumi/pulumi-cloud#158]. The other half
involves changes in [pulumi/pulumi-terraform].
If a blob's reported size is incorrect, `archiveTar` may attempt to
write more bytes to an entry than it reported in that entry's header.
These changes provide a bit more context with the resulting error as
well as removing an unnecessary `LimitReader`.
Adds OpenTracing in the Pulumi engine and plugin + langhost subprocesses.
We currently create a single root span for any `Enging.plan` operation - which is a single `preview`, `update`, `destroy`, etc.
The only sub-spans we currently create are at gRPC boundaries, both on the client and server sides and on both the langhost and provider plugin interfaces.
We could extend this to include spans for any other semantically meaningful sections of compute inside the engine, though initial examples show we get pretty good granularity of coverage by focusing on the gRPC boundaries.
In the future, this should be easily extensible to HTTP boundaries and to track other bulky I/O like datastore read/writes once we hook up to the PPC and Pulumi Cloud.
We expose a `--trace <endpoint>` option to enable tracing on the CLI, which we will aim to thread through to subprocesses.
We currently support sending tracing data to a Zipkin-compatible endpoint. This has been validated with both Zipkin and Jaeger UIs.
We do not yet have any tracing inside the TypeScript side of the JS langhost RPC interface. There is not yet automatic gRPC OpenTracing instrumentation (though it looks like it's in progress now) - so we would need to manually create meaningful spans on that side of the interface.
We currently have a nasty issue with archive assets wherein they read
their entire contents into memory each time they are accessed (e.g. for
hashing or translation). This interacts badly with scenarios that
place large amounts of data in an archive: aside from limiting the size
of an archive the engine can handle, it also bloats the engine's memory
requirements. This appears to have caused issues when running the PPC in
AWS: evidence suggests that the very high peak memory requirements this
approach implies caused high swap traffic that impacted the service's
availability.
In order to fix this issue, these changes move archives onto a
streaming read model. In order to read an archive, a user:
- Opens the archive with `Archive.Open`. This returns an ArchiveReader.
- Iterates over its contents using `ArchiveReader.Next`. Each returned
blob must be read in full between successive calls to
`ArchiveReader.Next`. This requirement is essentially forced upon us
by the streaming nature of TAR archives.
- Closes the ArchiveReader with `ArchiveReader.Close`.
This model does not require that the complete contents of the archive or
any of its constituent files are in memory at any given time.
Fixes#325.
Rework the polling code for `pulumi` when targeting hosted scenarios. (i.e. absorb https://github.com/pulumi/pulumi-service/pull/282.) We now return an `updateID` for update, preview, and destroy. And use bake that into the URL.
This PR also silently ignores 504 errors which are now returned from the Pulumi Service if the PPC request times out.
Combined, this should ameliorate some of the symptoms we see from https://github.com/pulumi/pulumi-ppc/issues/60. Though we'll continue to try to identify and fix the root cause.
We are not ready to start running integration tests against
Pulumi.com, but `pulumi` uses `PULUMI_API` to decide what backend to
target. Remove it when running `pulumi` in the integration tests so we
always get the local behavior.
pulumi/pulumi#471 tracks the work to actually be able to choose what
backend you want when an integration test runs.
This changes the overall flow to exit early when a failure occurs.
We will try to clean up if we believe to have gotten far enough to
actually have provisioned any resources.
In our existing code, we only use the input state for old and new
properties. This is incorrect and I'm astonished we've been flying
blind for so long here. Some resources require the output properties
from the prior operation in order to perform updates. Interestingly,
we did correclty use the full synthesized state during deletes.
I ran into this with the AWS Cloudfront Distribution resource,
which requires the etag from the prior operation in order to
successfully apply any subsequent operations.
We were previously calling configure on each package once per time it was mentioned in the config. We only need to call it once ever as we pass the full bag of relevent config through on that one call.
It's legal and possible for undefined properties to show up in
objects, since that's an idiomatic JavaScript way of initializing
missing properties. Instead of failing for these during deployment,
we should simply skip marshaling them to Terraform and let it do
its thing as usual. This came up during our customer workload.
Previously, we stored configuration information in the Pulumi.yaml
file. This was a change from the old model where configuration was
stored in a special section of the checkpoint file.
While doing things this way has some upsides with being able to flow
configuration changes with your source code (e.g. fixed values for a
production stack that version with the code) it caused some friction
for the local development scinerio. In this case, setting
configuration values would pend changes to Pulumi.yaml and if you
didn't want to publish these changes, you'd have to remember to remove
them before commiting. It also was problematic for our examples, where
it was not clear if we wanted to actually include values like
`aws:config:region` in our samples. Finally, we found that for our
own pulumi service, we'd have values that would differ across each
individual dev stack, and publishing these values to a global
Pulumi.yaml file would just be adding noise to things.
We now adopt a hybrid model, where by default configuration is stored
locally, in the workspace's settings per project. A new flag `--save`
tests commands to actual operate on the configuration information
stored in Pulumi.yaml.
With the following change, we have have four "slots" configuration
values can end up in:
1. In the Pulumi.yaml file, applies to all stacks
2. In the Pulumi.yaml file, applied to a specific stack
3. In the local workspace.json file, applied to all stacks
4. In the local workspace.json file, applied to a specific stack
When computing the configuration information for a stack, we apply
configuration in the above order, overriding values as we go
along.
We also invert the default behavior of the `pulumi config` commands so
they operate on a specific stack (i.e. how they did before
e3610989). If you want to apply configuration to all stacks, `--all`
can be passed to any configuration command.
Add `Name` (Pulumi project name) and `Runtime` (Pulumi runtime, e.g. "nodejs") properties to `UpdateProgramRequest`; as they are now required.
The long story is that as part of the PPC enabling destroy operations, data that was previously obtained from `Pulumi.yaml` is now required as part of the update request. This PR simply provides that data from the CLI.
This is the final step of a line of breaking changes.
pulumi-ppc 8ddce15b29
pulumi-service 8941431d57 (diff-05a07bc54b30a35b10afe9674747fe53)
On windows, we have to indirect through a batch file to launch plugins,
which means when we go to close a plugin, we only kill cmd.exe that is
running the batch file and not the underlying node process. This
prevents `pulumi` from exiting cleanly. So on Windows, we also kill any
direct children of the plugin process
Fixes#504
This PR adds integration tests for exercising `pulumi init` and the `pulumi stack *` commands. The only functional change is merging in https://github.com/pulumi/pulumi/pull/492 , which I found while writing the tests and (of course 😁 ) wrote a regression for.
To do this I introduce a new test driver called `PulumiProgram`. This is different from the one found in the `testing/integration`package in that it doesn't try to prescribe a workflow. It really just deals in executing commands, and confirming strings are in the output.
While it doesn't hurt to have more tests for `pulumi`, my motivation here was so that I could reuse these to ensure I keep the same behavior for my pending PR that implements Cloud-enabled variants of some of these commands.
- When looking for a `.pulumi` folder to reuse, only consider ones
that have a `settings.json` in them.
- Make the error message when there is no repository a little more
informative by telling someone to run `pulumi init`
Calls to `NewProjectWorkspace` will panic if there is no Pulumi.yaml found in the folder hierarchy. Simple repo:
```
git init
pulumi init
pulumi stack init x
```
`DetectPackage` will search until it gets to the `.pulumi` directory or finds a `Pulumi.yaml` file. In the case of the former, we pass "" to `LoadPackage` which then asserts because the file doesn't exist.
We now detect this condition and surface an error to the user. The error text is patterned after running a git command when there is no .git folder found:
fatal: Not a git repository (or any of the parent directories): .git
This adds a minimal runtime verification test to our basic
test suite, to at least exercise the portions of the integration
test library that load up and parse checkpoint files.
Now, instead of having a .pulumi folder next to each project, we have
a single .pulumi folder in the root of the repository. This is created
by running `pulumi init`.
When run in a git repository, `pulumi init` will place the .pulumi
file next to the .git folder, so it can be shared across all projects
in a repository. When not in a git repository, it will be created in
the current working directory.
We also start tracking information about the repository itself, in a
new `repo.json` file stored in the root of the .pulumi folder. The
information we track are "owner" and "name" which map to information
we use on pulumi.com.
When run in a git repository with a remote named origin pointing to a
GitHub project, we compute the owner and name by deconstructing
information from the remote's URL. Otherwise, we just use the current
user's username and the name of the current working directory as the
owner and name, respectively.
This PR enables the `pulumi stack init` to work against the Pulumi Cloud. Of note, I using the approach described in https://github.com/pulumi/pulumi-service/issues/240. The command takes an optional `--cloud` parameter, but otherwise will use the "default cloud" for the target organization.
I also went back and revised `pulumi update` to do this as well. (Removing the `--cloud` parameter.)
Note that neither of the commands will not work against `pulumi-service` head, as they require some API refactorings I'm working on right now.
I noticed in our Docker builds, we often end up seeing %(MISSING)!
style messages, which were an indication we were trying to format
them. The reason was the presence of %c's in the stream, and the
fact that we passed said messages to Fprintf. We were careful in
all other layers to use the message on the "right hand side" of
any *f calls, but in this instance, we used Fprintf and passed the
message on the "left hand side", triggering formatting. It turns
out we've already formatted everything by the time we get here,
so there's no need -- we can just use Fprint instead.
Adds `pulumi update` so you can deploy to the Pulumi Console (via PPC on the backend).
As per an earlier discussion (now lost because I rebased/squashed the commits), we want to be more deliberate about how to bifurcate "local" and "cloud" versions of every Pulumi command.
We can block this PR until we do the refactoring to have `pulumi` commands go through a generic "PulumiCloud" interface. But it would be nice to commit this so I can do more refining of the `pulumi` -> Console -> PPC workflow.
Another known area that will need to be revisited is how we render the PPC events on the CLI. Update events from the PPC are generated in a different format than the `engine.Event`, and we'll probably want to change the PPC to emit messages in the same format. (e.g. how we handle coloring, etc.)
We now encrypt secrets at rest based on a key derived from a user
suplied passphrase.
The system is designed in a way such that we should be able to have a
different decrypter (either using a local key or some remote service
in the Pulumi.com case in the future).
Care is taken to ensure that we do not leak decrypted secrets into the
"info" section of the checkpoint file (since we currently store the
config there).
In addtion, secrets are "pay for play", a passphrase is only needed
when dealing with a value that's encrypted. If secure config values
are not used, `pulumi` will never prompt you for a
passphrase. Otherwise, we only prompt if we know we are going to need
to decrypt the value. For example, `pulumi config <key>` only prompts
if `<key>` is encrypted and `pulumi deploy` and friends only prompt if
you are targeting a stack that has secure configuration assoicated
with it.
Secure values show up as unecrypted config values inside the language
hosts and providers.
Trapping these signals hijacks the usual termination behavior for any
program that happens to link in the engine and perform an operation
that starts a gRPC server. These servers already provide a cancellation
mechanism via a `cancel` channel parameter; if the using program wants
to gracefully terminate these servers on some signal, it is responsible
for providing that behavior.
This also fixes a leak in which the goroutine responsible for waiting on
a server's signals and cancellation channel would never exit.
When `PULUMI_RETAIN_CHECKPOINTS` is set in the environment, also write
the checkpoint file to <stack-name>.<ext>.<timestamp>.
This ensures we have historical information about every snapshot, which
would aid in debugging issues like #451. We set this to true for our
integration tests.
Fixes#453
This change fixes a couple bugs with assets:
* We weren't recursing into subdirectories in the new "path as
archive" feature, which meant we missed most of the files.
* We need to make paths relative to the root of the archive
directory itself, otherwise paths end up redundantly including
the asset's root folder path.
* We need to clean the file paths before adding them to the
archive asset map, otherwise they are inconsistent between the
path, tar, tgz, and zip cases.
* Ignore directories when traversing zips, since they aren't
included in the other formats.
* Tolerate io.EOF errors when reading the ZIP contents into blobs.
* Add test cases for the four different archive kinds.
This fixespulumi/pulumi-aws#50.
The prior code was a little too aggressive in rejected undefined
properties, because it assumed any occurrence indicated a resource
that was unavailable due to planning. This is a by-produt of our
relatively recent decision to flow undefineds freely during planning.
The problem is, it's entirely legitimate to have undefined values
deep down in JavaScript structures, entirely unrelated to resources
whose property values are unknown due to planning.
This change flows undefined more freely. There really are no
negative consequences of doing so, and avoids hitting some overly
aggressive assertion failures in some important scenarios. Ideally
we would have a way to know statically whether something is a resource
property, and tighten up the assertions just to catch possible bugs
in the system, but because this is JavaScript, and all the assertions
are happening at runtime, we simply lack the necessary metadata to do so.
Now that config is stored in Pulumi.yaml, we need to mimic the behavior
around .pulumi/ while edits are applied. This will ensure that config
values carry forward from the original program settings.
This fixespulumi/pulumi-aws#48.
The event diagnostic goroutines could error out sometimes during
early program exits, due to a race between the goroutine writing to
the channel and the early exiting goroutine which closed the channel.
This change stops closing the channels entirely on the abrupt exit
paths, since it's not necessary and we want to exit immediately.
This improves a few things about assets:
* Compute and store hashes as input properties, so that changes on
disk are recognized and trigger updates (pulumi/pulumi#153).
* Issue explicit and prompt diagnostics when an asset is missing or
of an unexpected kind, rather than failing late (pulumi/pulumi#156).
* Permit raw directories to be passed as archives, in addition to
archive formats like tar, zip, etc. (pulumi/pulumi#240).
* Permit not only assets as elements of an archive's member list, but
also other archives themselves (pulumi/pulumi#280).
The existing `SnapshotProvider` interface does not sufficiently lend
itself to reliable persistence of snapshot data. For example, consider
the following:
- The deployment engine creates a resource
- The snapshot provider fails to save the updated snapshot
In this scenario, we have no mechanism by which we can discover that the
existing snapshot (if any) does not reflect the actual state of the
resources managed by the stack, and future updates may operate
incorrectly. To address this, these changes split snapshotting into two
phases: the `Begin` phase and the `End` phase. A provider that needs to
be robust against the scenario described above (or any other scenario
that allows for a mutation to the state of the stack that is not
persisted) can use the `Begin` phase to persist the fact that there are
outstanding mutations to the stack. It would then use the `End` phase to
persist the updated snapshot and indicate that the mutation is no longer
outstanding. These steps are somewhat analogous to the prepare and
commit phases of two-phase commit.
If --logtostderr is passed, and an unhandled error occurs that
was produced by the github.com/pkg/errors package, we will now
emit the stack trace. Much easier for debugging purposes.
The change to use a Goroutine for pumping output causes a hang
when an error occurs. This is because we unconditionally block
on the <-done channel, even though the failure means the done
will actually never occur. This changes the logic to only wait
on the channel if we successfully began the operation in question.
Instead of doing the logic to see if a type has YAML tags and then
dispatching based on that to use either the direct go-yaml marshaller
or the one that works in terms of JSON tags, let's just say that we
always add YAML tags as well, and use go-yaml directly.
Previously, config information was stored per stack. With this change,
we now allow config values which apply to every stack a program may
target.
When passed without the `-s <stack>` argument, `pulumi config`
operates on the "global" configuration. Stack specific information can
be modified by passing an explicit stack.
Stack specific configuration overwrites global configuration.
Conside the following Pulumi.yaml:
```
name: hello-world
runtime: nodejs
description: a hello world program
config:
hello-world:config:message Hello, from Pulumi
stacks:
production:
config:
hello-world:config:message Hello, from Production
```
This program contains a single configuration value,
"hello-world:config:message" which has the value "Hello, from Pulumi"
when the program is activated into any stack except for "production"
where the value is "Hello, from Production".
Instead of having information stored in the checkpoint file, save it
in the Pulumi.yaml file. We introduce a new section `stacks` which
holds information specific to a stack.
Next, we'll support adding configuration information that applies
to *all* stacks for a Program and allow the stack specific config to
overwrite or augment it.
This will allow us to marhsall and unmarshall the structure using the
go-yaml package directly, instead of the current way we handle YAML
which is to treat it as JSON and use the JSON marshaller behind the
scenes.
By having the explicit tags and using go-yaml directly, we can ensure
the resulting order of the resulting document's properties matches the
go type instead of just being lexigraphicaly sorted.
Long term, we probably want to stop using go-yaml or other packages in
favor of parsing the YAML file into some sort of DOM that we can use
to retain comments and other formatting in the file.
This PR adds `login` and `logout` commands to the `pulumi` CLI.
Rather than requiring a user name and password like before, we instead require users to login with GitHub credentials on the Pulumi Console website. (You can do this now via https://beta.moolumi.io.) Once there, the account page will show you an "access token" you can use to authenticate against the CLI.
Upon successful login, the user's credentials will be stored in `~/.pulumi/credentials.json`. This credentials file will be automatically read with the credentials added to every call to `PulumiRESTCall`.
This change adds functions, `pulumi.getProject()` and `pulumi.getStack()`,
to fetch the names of the project and stack, respectively. These can be
handy in generating names, specializing areas of the code, etc.
This fixespulumi/pulumi#429.
During the course of a `pulumi update`, it is possible for a resource to
become slated for deletion. In the case that this deletion is part of a
replacement, another resource with the same URN as the to-be-deleted
resource will have been created earlier. If the `update` fails after the
replacement resource is created but before the original resource has been
deleted, the snapshot must capture that the original resource still exists
and should be deleted in a future update without losing track of the order
in which the deletion must occur relative to other deletes. Currently, we
are unable to track this information because the our checkpoints require
that no two resources have the same URN.
To fix this, these changes introduce to the update engine the notion of a
resource that is pending deletion and change checkpoint serialization to
use an array of resources rather than a map. The meaning of the former is
straightforward: a resource that is pending deletion should be deleted
during the next update.
This is a fairly major breaking change to our checkpoint files, as the
map of resources is no more. Happily, though, it makes our checkpoint
files a bit more "obvious" to any tooling that might want to grovel
or rewrite them.
Fixes#432, #387.
A dynamic resource is a resource whose provider is implemented alongside
the resource itself. This provider may close over and use orther
resources in the implementation of its CRUD operations. The provider
itself must be stateless, as each CRUD operation for a particular
dynamic resource type may use an independent instance of the provider.
Changes to the definition of a resource's provider result in replacement
of the resource itself (rather than a simple update), as this allows the
old provider definition to delete the old resource and the new provider
definition to create an appropriate replacement.
If a plugin fails to load after we've set up the goroutines that copy
from its std{out,err} streams, then those goroutines can end up writing
to a closed event channel. This change ensures that we properly drain
those streams in this case.
In #411 we started to run tests in parallel again. To support that, we
added a .yarnrc file in the root of our repository to pass --mutex
network to all yarn invocations, because tests may run yarn commands
concurrently with one another and yarn is not safe to run
concurrently.
However, when we run the integration tests, we actually copy them into
a folder under `/tmp` and so yarn's logic to walk up the directory to
tree to find a `.yarnrc` will not see this `.yarnrc` and we're back
where we started.
Have the testing package explicitly write this file, which should
prevent this issue from happening in the future.
Previously we used the word "Environment" as the term for a deployment
target, but since then we've started to use the term Stack. Adopt this
across the CLI.
From a user's point of view, there are a few changes:
1. The `env` verb has been renamed to `stack`
2. The `-e` and `--env` options to commands which operate on an
environment now take `-s` or `--stack` instead.
3. Becase of (2), the commands that used `-s` to display a summary now
only support passing the full option name (`--summary`).
On the local file system, we still store checkpoint data in the `env`
sub-folder under `.pulumi` (so we can reuse existing checkpoint files
that were written to the old folder)
This changes a few things about "components":
* Rename what was previously ExternalResource to CustomResource,
and all of the related fields and parameters that this implies.
This just seems like a much nicer and expected name for what
these represent. I realize I am stealing a name we had thought
about using elsewhere, but this seems like an appropriate use.
* Introduce ComponentResource, to make initializing resources
that merely aggregate other resources easier to do correctly.
* Add a withParent and parentScope concept to Resource, to make
allocating children less error-prone. Now there's no need to
explicitly adopt children as they are allocated; instead, any
children allocated as part of the withParent callback will
auto-parent to the resource provided. This is used by
ComponentResource's initialization function to make initialization
easier, including the distinction between inputs and outputs.
This change implements core support for "components" in the Pulumi
Fabric. This work is described further in pulumi/pulumi#340, where
we are still discussing some of the finer points.
In a nutshell, resources no longer imply external providers. It's
entirely possible to have a resource that logically represents
something but without having a physical manifestation that needs to
be tracked and managed by our typical CRUD operations.
For example, the aws/serverless/Function helper is one such type.
It aggregates Lambda-related resources and exposes a nice interface.
All of the Pulumi Cloud Framework resources are also examples.
To indicate that a resource does participate in the usual CRUD resource
provider, it simply derives from ExternalResource instead of Resource.
All resources now have the ability to adopt children. This is purely
a metadata/tagging thing, and will help us roll up displays, provide
attribution to the developer, and even hide aspects of the resource
graph as appropriate (e.g., when they are implementation details).
Our use of this capability is ultra limited right now; in fact, the
only place we display children is in the CLI output. For instance:
+ aws:serverless:Function: (create)
[urn=urn:pulumi:demo::serverless::aws:serverless:Function::mylambda]
=> urn:pulumi:demo::serverless::aws:iam/role:Role::mylambda-iamrole
=> urn:pulumi:demo::serverless::aws:iam/rolePolicyAttachment:RolePolicyAttachment::mylambda-iampolicy-0
=> urn:pulumi:demo::serverless::aws:lambda/function:Function::mylambda
The bit indicating whether a resource is external or not is tracked
in the resulting checkpoint file, along with any of its children.
Previously, the engine would write to io.Writer's to display output.
When hosted in `pulumi` these writers were tied to os.Stdout and
os.Stderr, but other applications hosting the engine could send them
other places (e.g. a log to be sent to an another application later).
While much better than just using the ambient streams, this was still
not the best. It would be ideal if the engine could just emit strongly
typed events and whatever is hosting the engine could care about
displaying them.
As a first step down that road, we move to a model where operations on
the engine now take a `chan engine.Event` and during the course of the
operation, events are written to this channel. It is the
responsibility of the caller of the method to read from the channel
until it is closed (singifying that the operation is complete).
The events we do emit are still intermingle presentation with data,
which is unfortunate, but can be improved over time. Most of the
events today are just colorized in the client and printed to stdout or
stderr without much thought.
The checkpoint is an implementation detail of the storage of an
environment. Instead of interacting with it, make sure that all the
data we need from it either hangs off the Snapshot or Target
objects (which you can get from a Checkpoint) and then start consuming
that data.
This let's you set things like YARNFLAGS==--offline which is helpful
when you are on an airplane. Yarn can still pick up stuff that you had
pulled down recently from its local cache
This change adds the capability for a resource provider to indicate
that, where an action carried out in response to a diff, a certain set
of properties would be "stable"; that is to say, they are guaranteed
not to change. As a result, properties may be resolved to their final
values during previewing, avoiding erroneous cascading impacts.
This avoids the ever-annoying situation I keep running into when demoing:
when adding or removing an ingress rule to a security group, we ripple
the impact through the instance, and claim it must be replaced, because
that instance depends on the security group via its name. Well, the name
is a great example of a stable property, in that it will never change, and
so this is truly unfortunate and always adds uncertainty into the demos.
Particularly since the actual update doesn't need to perform replacements.
This resolvespulumi/pulumi#330.
Now that `deploy.Step.Pre` is no more, we can simplify the `StepActions`
interface down to a single method, `Run`, which performs all actions
associated with the step. This feeds into #388.
Previously, the engine was concered with maintaing information about
the currently active environment. Now, the CLI is in charge of
this. As part of this change, the engine can now assume that every
environment has a non empty name (and I've added asserts on the
entrypoints of the engine API to ensure that any consumer of the
engine passes a non empty environment name)
Simply write the current snapshot after each step. We should probably be
more clever about this in the future (i.e. write only the changes rather
than the entire file).
It was only being used for two cases where we would issue warnings for
cases where the file system casing did not match expected casing. I
think it's probably better if we don't try to be smart here and just
treat these cases the same as if the file had not existed. Removing
the dependncy on diag also makes it a little clearer that this stuff
should be pulled out from the engine.
`deploy.Plan.Apply` was only consumed by the engine, and seemed to be in
the wrong place given the API exported by the rest of `Plan` (i.e.
`Plan.Start` + `PlanIterator`). Furthermore, we were missing a reasonable
opportunity to share code between `update` and `preview`, both of which
need to walk the plan. These changes move the plan walk into `package engine`
as `planResult.Walk` and replace the `Progress` interface with a new interface,
`StepActions`, which subsumes the functionality of the former and adds support
for implementation-specific step execution. `planResult.Walk` is then
consumed by both `Engine.Deploy` and `Engine.PrintPlan`.
The CLI is now responsible for actually displaying information and the
engine is only concerned with getting the configuration. As part of
this change, I've removed the display a single configuration value API
from the engine. It can now be done in terms of getting all the config
for an environment and selecting the value the user is interested in
Previously the engine was concerned with displaying information about
the environment. Now the engine returns an environment info object
which the CLI uses to display environment information.
This change enables us to make progress on exposing data sources
(see pulumi/pulumi-terraform#29). The idea is to have an Invoke
function that simply takes a function token and arguments, performs
the function lookup and invocation, and then returns a return value.
Print "modified" rather than "modifyd". This introduces a new method,
`resource.StepOp.PastTense()`, which returns the past tense description
of the operation.
The current `pulumi env` command just prints the deployment's info map
as-is, which leads to some ugly Go internal map printing output. Instead
of doing that, let's show a JSON-like map that is properly pretty printed.
This change improves our output formatting by generally adding
fewer prefixes. As shown in pulumi/pulumi#359, we were being
excessively verbose in many places, including prefixing every
console.out with "langhost[nodejs].stdout: ", displaying full
stack traces for simple errors like missing configuration, etc.
Overall, this change includes the following:
* Don't prefix stdout and stderr output from the program, other
than the standard "info:" prefix. I experimented with various
schemes here, but they all felt gratuitous. Simply emitting
the output seems fine, especially as it's closer to what would
happen if you just ran the program under node.
* Do NOT make writes to stderr fail the plan/deploy. Previously
we assumed that any console.errors, for instance, meant that
the overall program should fail. This simply isn't how stderr
is treated generally and meant you couldn't use certain
logging techniques and libraries, among other things.
* Do make sure that stderr writes in the program end up going to
stderr in the Pulumi CLI output, however, so that redirection
works as it should. This required a new Infoerr log level.
* Make a small fix to the planning logic so we don't attempt to
print the summary if an error occurs.
* Finally, add a new error type, RunError, that when thrown and
uncaught does not result in a full stack trace being printed.
Anyone can use this, however, we currently use it for config
errors so that we can terminate with a pretty error message,
rather than the monstrosity shown in pulumi/pulumi#359.
This includes a few changes:
* The repo name -- and hence the Go modules -- changes from pulumi-fabric to pulumi.
* The Node.js SDK package changes from @pulumi/pulumi-fabric to just pulumi.
* The CLI is renamed from lumi to pulumi.
While these codes can be understood on newer versions of Windows 10 (if
we were to set the console properties correctly) in general they do not
work and cause garbage to be printed to the screen. For now, just don't
colorize output on Windows.
Instead of binding on 0.0.0.0 (which will listen on every interface)
let's only listen on localhost. On windows, this both makes the
connection Just Work and also prevents the Windows Firewall from
blocking the listen (and displaying UI saying it has blocked an
application and asking if the user should allow it)
This change flips the polarity on parallelism: rather than having a
--serialize flag, we will have a --parallel=P flag, and by default
we will shut off parallelism. We aren't benefiting from it at the
moment (until we implement pulumi/pulumi-fabric#106), and there are
more hidden dependencies in places like AWS Lambdas and Permissions
than I had realized. We may revisit the default, but this allows
us to bite off the messiness of dependsOn only when we benefit from
it. And in any case, the --parallel=P capability will be useful.
This change adds an optiona dependsOn parameter to Resource constructors,
to "force" a fake dependency between resources. We have an extremely strong
desire to resort to using this only in unusual cases -- and instead rely
on the natural dependency DAG based on properties -- but experience in other
resource provisioning frameworks tells us that we're likely to need this in
the general case. Indeed, we've already encountered the need in AWS's
API Gateway resources... and I suspect we'll run into more especially as we
tackle non-serverless resources like EC2 Instances, where "ambient"
dependencies are far more commonplace.
This also makes parallelism the default mode of operation, and we have a
new --serialize flag that can be used to suppress this default behavior.
Full disclosure: I expect this to become more Make-like, i.e. -j 8, where
you can specify the precise width of parallelism, when we tackle
pulumi/pulumi-fabric#106. I also think there's a good chance we will flip
the default, so that serial execution is the default, so that developers
who don't benefit from the parallelism don't need to worry about dependsOn
in awkward ways. This tends to be the way most tools (like Make) operate.
This fixespulumi/pulumi-fabric#335.
This change finishes the conversion of LUMIDL over to the new
runtime model, with the appropriate code generation changes.
It turns out the old model actually had a flaw in it anyway that we
simply didn't hit because we hadn't been stressing output properties
nearly as much as the new model does. This resulted in needing to
plumb the rejection (or allowance) of computed properties more
deeply into the resource property marshaling/unmarshaling logic.
As of these changes, I can run the GitHub provider again locally.
This change fixespulumi/pulumi-fabric#332.
* Initialize the diganostics logger with opts.Debug when doing
a Deploy, like we do Plan.
* Don't spew leaked promises if there were Log.errors.
* Serialize logging RPC calls so that they can't appear out of order.
* Print stack traces in more places and, in particular, remember
the original context for any errors that may occur asynchronously,
like resource registration and calls to mapValue.
* Include origin stack traces generally in more error messages.
* Add some more mapValue test cases.
* Only undefined-propagate mapValue values during dry-runs.
This change upgrades gRPC to 1.6.0 to pick up a few bug fixes.
We also use the full address for gRPC endpoints, including the
interface name, as otherwise we pick the wrong interface on Linux.
There's a fair bit of clean up in here, but the meat is:
* Allocate the language runtime gRPC client connection on the
goroutine that will use it; this eliminates race conditions.
* The biggie: there *appears* to be a bug in gRPC's implementation
on Linux, where it doesn't implement WaitForReady properly. The
behavior I'm observing is that RPC calls will not retry as they
are supposed to, but will instead spuriously fail during the RPC
startup. To work around this, I've added manual retry logic in
the shared plugin creation function so that we won't even try
to use the client connection until it is in a well-known state.
pulumi/pulumi-fabric#337 tracks getting to the bottom of this and,
ideally, removing the work around.
The other minor things are:
* Separate run.js into its own module, so it doesn't include
index.js and do a bunch of random stuff it shouldn't be doing.
* Allow run.js to be invoked without a --monitor. This makes
testing just the run part of invocation easier (including
config, which turned out to be super useful as I was debugging).
* Tidy up some messages.
* Use `global.hasOwnProperty(ident)`, rather than `global[ident] !== undefined`,
to avoid classifying references to globals as free variables. Surprise(!!),
the prior logic wouldn't work for `undefined` itself... 😒
* Expand this check to include the built-in Node.js module variables, namely
`__dirname`, `__filename`, `exports`, `module`, and `require`, so that
references to them don't get classified as serializable free variables either.
* Place catch variables in scope, so that `catch (err) { ... }` won't yield
free variables for references to `err` within `...`.
* Place recursive function definitions into the top-level `var`-like scope of
variables so that we don't consider references to them free.
* Harden all error pathways in the native C++ add-on so that we terminate
anytime an exception is in-flight, rather than limping along and making
things worse...
If a resource's planning operation is to do nothing, we can safely
assume that all of its properties are stable. This can be used during
planning to avoid cascading updates that we know will never happen.
This change adds the ability to perform runtime logging, including
debug logging, that wires up to the Pulumi Fabric engine in the usual
ways. Most stdout/stderr will automatically go to the right place,
but this lets us add some debug tracing in the implementation of the
runtime itself (and should come in handy in other places, like perhaps
the Pulumi Framework and even low-level end-user code).
As explained in pulumi/pulumi-fabric#293, we were a little ad-hoc in
how configuration was "applied" to resource providers.
In fact, config wasn't ever communicated directly to providers; instead,
the resource providers would simply ask the engine to read random heap
locations (via tokens). Now that we're on a plan where configuration gets
handed to the program at startup, and that's that, and where generally
speaking resource providers never communicate directly with the language
runtime, we need to take a different approach.
As such, the resource provider interface now offers a Configure RPC
method that the resource planning engine will invoke at the right
times with the right subset of configuration variables filtered to
just that provider's package. This fixespulumi/pulumi#293.
This change simplifies the provider RPC interface slightly:
1) Eliminate Get. We really don't need it anymore. There are
several possibly-interesting scenarios down the road that may
demand it, but when we get there, we can consider how best to
bring this back. Furthermore, the old-style Get remains mostly
incompatible with Terraform anyway.
2) Pass URNs, not type tokens, across the RPC boundary. This gives
the provider access to more interesting information: the type,
still, but also the name (which is no longer an object property).
This makes a few tweaks to get the integration tests passing:
* Add `runtime: nodejs` to the minimal example's `Lumi.yaml` file.
* Remove usage of `@lumi/lumirt { printf }` and just use `console.log`.
* Remove calls to `lumijs` in the integration test framework and
the minimal example's package.json. Instead, we just run
`yarn run build`, which itself internally just invokes `tsc`.
* Add package validation logic and eliminate the pkg/compiler/metadata
library, in favor of the simpler code in pkg/engine.
* Simplify the Node.js langhost plugin CLI, and simply take an
argument rather than requiring required and optional --flags.
* Use a default path of "." if the program path isn't provided. This
is a legal scenario if you've passed a pwd and just want to load
the package's default module ("./index.js" or whatever main says).
* Add an executable script, lumi-langhost-nodejs, that fires up the
`bin/cmd/langhost/index.js` file to serve the Node.js language plugin.
The existing implementation of the interface (backed by the file
system) has moved into cmd/lumi. The deployment service will start to
provide its own version.
`saveEnv` had a flag which would prevent an environment from being
overwritten if it already existed, which was only used by `lumi env
init`. Refactor the code so the check is done inside `lumi` instead of
against this API. We don't need this functionality for the service and
so requiring support for this at the API boundary for environments
feels like a bad idea.
We'd like to abstract out environment CRUD operations and I'd prefer
not to have to bake in the conspect of a file name like thing in the
abstraction. Since we were not really using this feature many places,
let's just get rid of it.
The implementation of these functions will be moving out of the engine
and into `lumi` itself, it's a little easier if we move away from
spewing stuff to the diag interface, so just use glog instead (which
`lumi` already uses for logging)
This helps avoid running into file handle limits when creating archives including thousands of node_modules files.
Tracking a more complete fix through all other codepaths related to assets as part of #325.
This change ensures we close all Blobs in the asset/archive logic.
In particular, the archive.Read function returns a map of files to
Blobs and after we are done copying the contents we must ensure
that we invoke Close, otherwise we may leak file handles, sockets,
and so on. This may or may not be the culprit to the "too many
files open" errors we are hitting while deploying the M5 bits.
This refactors the engine so all of the APIs on it are instance
methods on the type instead of raw methods that float around and use
data from a global engine.
A mechcanical change as we remove the global `E` and then make
anything that interacted with that in pkg/engine to be an instance
method and the dealing with the fallout.
Instead of talking directly to stdout or stderr via methods on fmt,
indirect through an Engine type (presently a global, but soon to
change) to allow control of where the streams actually end up.
Prevously, we would throw raw args arrays across the interface and the
engine would do some additional parsing. Clean this up so we don't do
that and all the parsing stays in `lumi`
Previous logic had assumed arrays were treated like objects
in the runtime (like in ECMAScript), but they are a unique
value kind in current Lumi runtime, so must be handled separately.
Adds an `ExtraRuntimeValidation` hook to the test harness.
This runs after the test app is deployed, and can be used to test publically
exposed endpoints on the example to validate additional runtime correctness
of the Lumi app under test.
This changes a few things in the CLI, mostly just prettying it up:
* Label all steps more clearly with the kind of step. Also
unify the way we present this during planning and deployment.
* Summarize the changes that *did not* get made just as clearly
as those that did. In other words, stuff like this:
info: 2 resources changed:
+1 resource created
-1 resource deleted
5 resources unchanged
and
info: no resources required
5 resources unchanged
* Always print output properties when they are pertinent.
This includes creates, replacements, and updates.
* Show replacement creates and deletes very distinctly. The
create parts show up minty green and the delete parts show up
rosey red. These are the "physical" steps, compared to the
"logical" step of replacement (which remains marigold).
I still don't love where we are here. The asymmetry between
planning and deployment bugs me, and could be surprising.
("Hey, my deploy doesn't look like my plan!") I don't know
what developers will want to see here and I feel like in
general we are spewing far too much into the CLI to make it
even useful for anything but diagnosing failures afterwards.
I propose that we should do a deep dive on this during the
CLI epic, pulumi/pulumi-service#2.
This resolvespulumi/pulumi-fabric#305.
This change tests that a plan and deploy immediately following another
deploy, when no edits have taken place, correctly results in no action.
I also cleaned up a few things in the code, like using fmt.Printf rather
than fmt.Fprintf(os.Stdout, ...), to clean up error paths, giving the
package a slightly shorter name, and adding missing copyright headers.
This is part of pulumi/pulumi-fabric#310.
At the moment, we permit resources to carry a name, which the
engine uses as part of URN creation. Unfortunately, the property
"name" has a very high chance of conflicting with meaningful
user-authored properties. And furthermore this sort of name,
although key to creating URNs which are core to how the overall
system performs deployments and manages resources, are seldom
used programmatically in Lumi programs. As a result, it's a real
nuisance that we stole the good name.
This change renames that property from "name" to "urnName". This
not only has a lower likelihood of conflicting, but it also looks
reasonable sitting alongside the "urn" property. In fact, in some
future universe, after some of the upcoming runtime changes, we
may not even need the name property on the objects whatsoever.
I had originally toyed with the idea of eliminating the Resource
versus NamedResource distinction. This is certainly simpler and
remains a possibility. I didn't do that right now, however, because
the flexibility of letting resource providers name resources however
they see fit still seems possibly useful. For example, we keep
talking about whether functions can be auto-named based on hashes.
Until we've run those conversations to ground, I'd hate to do some
work that just needs to be undone in order to enable a scenario that
has a non-trivial likelihood of us wanting to explore.
We are now freely flowing computed and output properties across the
RPC boundary with providers. As such, we need to tolerate them in
a few more places. Namely, mapping to and from regular non-resource
property values, and also when copying RPC resource state back onto
live runtime objects.
If certain early errors occurred, like failing to find a default module
or main entrypoint, we never properly invoked OnDone (or, sometimes,
OnStart, for that matter). This meant that callers like the eval source
in the deployment engine could end up missing signals; in this particular
case, it led to a failure to signal a rendezvous synchronization object,
which itself led to a hang.
The fix is simple: make sure to call the On* methods in the right places.
I've added tests to probe the interesting paths, including failures.
This fixespulumi/pulumi-fabric#281.
We are renaming Lumi to Pulumi Fabric. This change simply renames the
pulumi/lumi repo to pulumi/pulumi-fabric, without the CLI tools and other
changes that will follow soon afterwards.
This changes the RPC interfaces between Lumi and provider ever so
slightly, so that we can track default properties explicitly. This
is required to perform accurate diffing between inputs provided by
the developer, inputs provided by the system, and outputs. This is
particularly important for default values that may be indeterminite,
such as those we use in the bridge to auto-generate unique IDs.
Otherwise, we fail to reapply defaults correctly, and trick the
provider into thinking that properties changed when they did not.
This is a small step towards pulumi/lumi#306, in which we will defer
even more responsibility for diffing semantics to the providers.
This change serializes unknown properties anywhere in the entire
property structure, including deeply embedded inside object maps, etc.
This is now done in such a way that we can recover both the computed
nature of the serialized property, along with its expected eventual
type, on the other side of the RPC boundary.
This will let us have perfect fidelity with the new bridge's view on
computed properties, rather than special casing them on "one side".
For Update and Delete operations, we provided just the input state
for a resource. This is insufficient, because the provider may need
to depend on output state from the Create or prior Update operations.
This change merges the output atop the input during the step application.
As part of the bridge bringup, I've discoverd that the property state
returned from Creates does *not* always equal the state that is then
read from calls to Get. (I suspect this is a bug and that they should
be equivalent, but I doubt it's fruitfal to try and track down all
occurrences of this; I bet it's widespread). To cope with this, we will
return state from Create and Update, instead of issuing a call to Get.
This was a design we considered to start with and frankly didn't have
a super strong reason to do it the current way, other than that it seemed
elegant to place all of the Get logic in one place.
Note that providers may choose to return nil, in which case we will read
state from the provider in the usual Get style.
This change mirrors the dynamic marshaling structure on the static
definition of the Asset and Archive types. This ensures that they
marshal correctly even when deeply embedded inside other structures.
This change brings the same typed serialization we use for RPC
to the serialization of deployments. This ensures that we get
repeatable diffs from one deployment to the next.
Until we support output-conditional code (pulumi/lumi#170), we
run into cases where we want to make a decision based on a computed
property if it is available, but can't possible know the value. In
such cases, we get an error ("Unexpected computed value").
In fact, our generated code currently includes client side validation
of properties, since it leads to a(n admittedly only slightly) better
developer experience. But that triggers the error for required
properties that are computed, basically hosing our ability to plan.
This change introduces a defaultIfComputed intrinsic to the lumirt
library that can be used to work around this. It takes two arguments,
obj and def, and returns the actual property value obj if it is NOT
computed; if it is computed, def will be substituted in its stead.
The generated code changes to use this. It's possible it will miss
validation, of course, if a computed property turns out not to match
the precondition. But this is strictly better than the alternative
of not trying to validate any of them to begin with.
This reverts commit c3db70849d.
I've opted to take a new strategy to ensure the bridge properties
don't conflict (with manual renames), similar to the name property.
This change recognizes assets and archives as 1st class resource
property values. This is necessary to support them in the new bridge
work, and lays the foundation for fixing pulumi/lumi#153.
I also took the opportunity to clean up some old cruft in the
resource properties area.
This renames the basemost resource properties, id and urn, to
names that are less likely to conflict with properties that real
resources will want to use, pid and upn (provider ID and Universal
Pulumi Name, respectively).
I actually ran into this with the current bridge work. An alternative
solution would be to require derived resources to pick different names,
however this is unfortunate because usually they are more "user-facing"
than ours. Another alternative is to not hijack the object properties
at all, but that too is problematic because we use these properties
during the evaluation of plans and deployments.
This seems like a reasonable middle ground.
This change introduces a --debug option to the plan, deploy, and
destroy commands. Unlike --logtostderr, which merely hooks into the
copious Glogging that we perform (and is therefore meant for developers
of the tools themselves and not end users), --debug hooks into the
user-facing debug stream. This now includes any debug messages coming
from the resource providers as they perform their tasks.
Remove duplicative call to UpdateFunctionConfiguration.
Also ensure that free variables returns stable variable order
to avoid unnecessary lambda updates.
Generalizes Lumi program validation so that it can be applied
to integration testing for other packages (such as the
pulumi/lumi-platform package examples).
Adds support for serializing lambdas which are
properties of captured objects.
For example:
```
let o = {
f: () => 12
};
let func = new aws.serverless.Function("func", {policies: [aws.iam.AWSLambdaFullAccess]}, (ev, ctx, cb) => {
console.log(o.f());
cb(null, null);
});
```
Generate names for serialized functions as hashes of the function text, and
bind these to the actual referenced names locally to each serialized closure.
This avoids potential conflicts between different functions which are referenced
with the same name.
For example:
```
function __8c176d768f95567f8c3acd08e486a1da7299fbc1() {
with({ f: __e61452ceb9d53f3611ed543f3822a85ad29345dd}) {
return (() => {
return function (cb) { cb(null, f(7)); };
})().apply(this, arguments);
}
}
```
Also fixes the stack overflows when there are cycles in the Closure graph by
adding caches of serialized functions.
Fixes#238.
This unblocks some cases with generics without having to
implement full generics support in the type LumiRT type system
(which we directionally will be removing anyway).
We would like to allow developers to use async/await
on the inside (Node.js) of Lumi programs.
We now support (don't error on) usage of async/await
inside runtime callbacks in Lumi programs. If await is
used during deployment, it will trigger an error.
Also adds support for try/catch in LumiJS, as these are
used more heavily in async/await code.
Since we target Node.js environments without native support
for async/await, we also emit runtime helpers to support TS
transpilation of async/await for Node.js pre-7.6.
This change rearranges some of the code generator logic in LumIDL
in such a way that we can use it from other code generators (like
the ongoing bridge work).
This adds a ReadLocations RPC function to the engine interface, alongside
the singular ReadLocation. The plural function takes a single token that
represents a module or class and we will then return all of the module
or class (static) properties that are currently known.
This adds a handy MapReplace function on pkg/resource's PropertyMap and
PropertyValue types. This is just like the existing Mappable function,
except that it permits easy replacement of elements as the map transformation
occurs. We need this to perform float64=>int transformations.
We fail very late in the process of plan application, should a duplicate
URN arise. This change fails as early in the process as possible and
ensures that it does so with good line number information.
This properly unwinds the interpreter should something happen that
results in cancellation. This occurs, for example, when the planning
engine encounters an error and decides that it doesn't need to proceed
further with evaluation before it simply goes ahead and exits.
The old contract library tried to be glog-friendly in its failfast behavior.
It turns out glog seldom does the right thing when goroutines are involved
(which, as of last sprint, they now are). We already had issues with stacks
not getting printed when --logtostderr was turned on, and the code tried
to work around this; but this still didn't work for the goroutines case.
All of this seems like way too much cleverness. Let's just use Go panics.
This change starts running Check with a "" property for cases where
global validation must take place (such as ensuring that required
configuration variables were set). It may be safely ignored if per-
property validation is preferred by a given resource.
Address several issues with running the Beanstalk
example in newer AWS regions with different requirements.
Ensures S3 bucket names adhere to required naming patterns
outside of us-east-1.
Also add InstanceProfile and ServiceRole configuration to the
beanstalk example as required in newer regions.
This adds a few missing closes for the plugin host/context. This
should fixpulumi/lumi#261. Eventually when we have more robust
nightly test options, and want to spend the time, we should think
about doing more rigorous stress testing that kills processes at
inopportune times and guarantees we don't leak. I've filed
pulumi/lumi#263 to do that.
This change fixes a few things:
* Most importantly, we need to place a leading "." in the paths
to Gometalinter, otherwise some sub-linters just silently skip
the directory altogether. errcheck is one such linter, which
is a very important one!
* Use an explicit Gometalinter.json file to configure the various
settings. This flips on a few additional linters that aren't
on by default (line line length checking). Sadly, a few that
I'd like to enable take waaaay too much time, so in the future
we may consider a nightly job (this includes code similarity,
unused parameters, unused functions, and others that generally
require global analysis).
* Now that we're running more, however, linting takes a while!
The core Lumi project now takes 26 seconds to lint on my laptop.
That's not terrible, but it's long enough that we don't want to
do the silly "run them twice" thing our Makefiles were previously
doing. Instead, we shall deploy some $$($${PIPESTATUS[1]}-1))-fu
to rely on the fact that grep returns 1 on "zero lines".
* Finally, fix the many issues that this turned up.
I think(?) we are done, except, of course, for needing to drive
down some of the cyclomatic complexity issues (which I'm possibly
going to punt on; see pulumi/lumi#259 for more details).
We were not propagating the error from `deployLatest` through
to the CLI error result. Despite out recent efforts to integrate
gometalinter, there were also several additional similar cases of
ignored error results reported by `errcheck`. Not yet clear why
these are not being reported via gometalinter.
Fixes#262.
After 233c5a8 landed, I noticed there are a few things to be fixed up:
* Run gometalinter in all the right places. We need to run both in
lint and lint_quiet targets. I've also cleaned up some of the logic
around what to suppress so there's less repetition.
* We currently @ meaningful commands, which is unfortunate, since it
makes debugging Makefiles tough (especially when looking at CI build
logs). Going forward, we should only use @ for meaningless commands,
like @echo.
* The AWS project wasn't actually running tslint, because it needs to
say `tslint './pack/**/*.ts' --exclude='./pack/node_modules/**'`.
The current script of `tslint lib/aws/pack/...` wasn't actually
running lint, hence we missed a lot of AWS lint issues.
* Fix up the issues that these fixes uncovered. Mostly err shadowing.
This continues the previous commit and establishes the interpreter
context so that we can use the new host interface. In summary:
* Instead of using the NullSource for destructions -- which
doesn't hook up an interpreter and so any reads of configuration
variables will fail -- we will enlighten the EvalSource to know
how to orchestrate destruction interpretation. The primary
difference is that we don't actually run the code, but *we do*
perform all of the necessary configuration and variable init.
* Associate the active interpreter with the plugin context as
we are executing, so that the host object can actually read the
state from the heap as requested to do so by attached plugins.
* Rename anything "engine" related to use the term "host"; this
avoids introducing unnecesarily new terminology.
* Add a new pkg/resource/provider/ package where we can begin
consolidating helper functionality for resource providers.
Right now, this includes a wrapper interface atop the gRPC
machinery necessary to contact the host, in addition to a
Main function that hides some boilerplate entrypoint code.
* Add a rpcutil.IsBenignCloseErr routine to let us ignore
"benign" gRPC errors that are knowingly returned at shutdown.
This commit completes pulumi/lumi#117.
This addresses CR feedback from @lukehoban; namely, that we should
be going through the Read API for location reads in the plugin host
to ensure that getters are invoked as appropriate.
I also made Location's various fields private so that we aren't
tempted to make this mistake elsewhere, effectively "forcing" us
to go through the accessor methods.
This change adds an engine gRPC interface, and associated implementation,
so that plugins may do interesting things that require "phoning home".
Previously, the engine would fire up plugins and talk to them directly,
but there was no way for a plugin to ask the engine to do anything.
The motivation here is so that plugins can read evaluator state, such
as config information, but this change also allows richer logging
functionality than previously possible. We will still auto-log any
stdout/stderr writes; however, explicit errors, warnings, informational,
and even debug messages may be written over the Log API.
We run tests in parallel and recently we began hitting a high enough
degree of parallelism that we've begun seeing "unsynchronized access
to map" errors in our test passes (intermittently). The root cause
is that access to the type symbol caches aren't synchronized. It would
be ideal if we actually rewired these to be cached in the compiler
context -- rather than being global -- but this fix is sufficient for
now. We will simply synchronize access using a Mutex.
The previous change broke the intrinsics tests in the
case that the `lumirt` package was not installed on
the system (which is the case in Travis given our build
order).
Reverting to the previous pattern of creating a fake
`lumirt` package to run the tests in to avoid the
dependency on an externally installed `lumirt`.
Implements correct QuoteJSONString behaviour per ECMAScript
spec.
Also refactors intrinsic test harness to make it easier to add new
intrinsics tests going forward.
This change implements the `get` function for resources. Per pulumi/lumi#83,
this allows Lumi scripts to actually read from the target environment.
For example, we can now look up a SecurityGroup from its ARN:
let group = aws.ec2.SecurityGroup.get(
"arn:aws:ec2:us-west-2:153052954103:security-group:sg-02150d79");
The returned object is a fully functional resource object. So, we can then
link it up with an EC2 instance, for example, in the usual ways:
let instance = new aws.ec2.Instance(..., {
securityGroups: [ group ],
});
This didn't require any changes to the RPC or provider model, since we
already implement the Get function.
There are a few loose ends; two are short term:
1) URNs are not rehydrated.
2) Query is not yet implemented.
One is mid-term:
3) We probably want a URN-based lookup function. But we will likely
wait until we tackle pulumi/lumi#109 before adding this.
And one is long term (and subtle):
4) These amount to I/O and are not repeatable! A change in the target
environment may cause a script to generate a different plan
intermittently. Most likely we want to apply a different kind of
deployment "policy" for such scripts. These are inching towards the
scripting model of pulumi/lumi#121, which is an entirely different
beast than the repeatable immutable infrastructure deployments.
Finally, it is worth noting that with this, we have some of the fundamental
underpinnings required to finally tackle "inference" (pulumi/lumi#142).
This change simplifies the generated Check interface for providers.
Instead of
Check(ctx context.Context, obj *T) ([]error, error)
where T is the resource type, we have
Check(ctx context.Context, obj *T, property string) error
This is done so that we can drive the calls to Check one property
at a time, allowing us to skip any that are computed. (Otherwise,
we may fail the verification erroneously.)
This has the added advantage that the Check implementations are
simpler and can simply return a single error. Furthermore, the
generated RPC code handles wrapping the result, so we can just do
return errors.New("bad");
rather than the previous reflection-laden junk
return resource.NewFieldError(
reflect.TypeOf(obj), awsservice.AWSResource_Property,
errors.New("bad"))
This change implements showing a summary of the current environment.
All you need to do is run
$ lumi env
and the current environment's information will be printed.
This makes it convenient to grab resource information that might be
required, for instance, to correlate with logs (e.g., lambda ARNs).
Eventually, as per pulumi/lumi#184, we want to print details about
all of the resources too.
Tests all of our commonly used examples.
Also sets test parallelism to 10 by default
since we are I/O bound on API calls to
the resource providers.
Also avoids using larger EC2 examples in
our samples so that we can keep our test
costs lower :-).
Adds an integration test that runs the following commands on the
AWS webserver example, failing if any command returns an error
code:
* lumijs
* lumi env init
* lumi config
* lumi plan
* lumi deploy
* lumi destroy
* lumi env rm
Also ensures that plan and deploy failures propagate errors through
to error codes at the CLI.
On the first turn, we want to distinguish between a coroutine
running that owns its turn, and a coroutine that knows it doesn't
own the turn and is simply awaiting its turn. The old Meet logic
wasn't quite right; instead, we'll have the caller tell us this.
This change fixes an issue with the way we deal with computed values in
assignments. Specifically, the assignment expression should resolve to
the computed value itself, but it must actually perform the assignment!
Previously, we evaluated to the right thing, but skipped the assignment.
We now have enough output properties implementation
working to change our API gateway examples and API
wrapper to correctly wire the API routes to the ARNs of
lambdas passed in to them.
We both wire up the lambda to the route, but also create
a permission specific to each route to assign to the
corresponding lambda - providing least privelege needed
for the API definition.
Also adds `string#toUpperCase` and fixes NewUniqueHex
to match how we are using it.
This implements array push and pop as intrinsics.
Also:
* Tighten up some assertions while I'm in here.
* Default initialize pointer slots to Null, if not done explicitly.
This change overwrites output property slots in runtime objects
after performing a CRUD operation, in addition to null or missing
slots, fixing #251. The problem is that we sometimes have output
property values pre-populated in an object, and sometimes don't,
depending on various things (both are legal). We should handle both.
The recent change to run the interpreter and planner on separate goroutines
created the need to perform rendezvous-style synchronization between them.
Although the case of an invoked function properly tore down the synchronization
by communicating the error, we seldom directly invoke functions for JavaScript
programs because the way module entrypoint code ends up in initializers.
This requires that we propagate errors correctly out of module and class
initializers, in the standard way, so that the unwind makes its way to the top.
This fixespulumi/lumi#246.
We recently changed the Resource base type to have no constructor,
rather than a manual empty constructor. This ought to work just fine.
The LumiJS compiler indeed generates a constructor, however, it is
missing a body and when the interpreter tries to invoke it, we crash
with a nil reference panic. The runtime actually tolerates missing
constructors entirely, although the way LumiJS binds super calls
doesn't tolerate the missing base constructor. This change simply
generates such constructors in LumiJS with empty bodies.
In addition, I've added an error that will catch the empty body
problem during binding, since technically speaking, all functions
must have bodies. (Our runtime happens to support the notion of
"abstract", however, so we only fire the error on concrete functions.)