Commit graph

2707 commits

Author SHA1 Message Date
Harshavardhana 00c697393a Merge pull request #1381 from minio/xl-layer
Implement XL layer
2016-04-25 15:59:30 -07:00
Harshavardhana 55032ffdf9 xl: Simplify blockingWriter and its usage. (#1373)
This removes odd races since we don't need to
track errors and avoids locking. All we need
is a Wait() and Done() waitgroup.
2016-04-25 12:47:31 -07:00
Harshavardhana 8bce699dae xl: Add logging. (#1372) 2016-04-25 12:47:31 -07:00
Harshavardhana 57f35c2bcc xl: Introduce new blocking writer to make CreateFile atomic. (#1362)
Creates a new write closer that must be released
by the read consumer. This is necessary so that
while commiting the underlying writers in erasure
coding we need to make sure we reply success only if
we have committed to disk.

This in turn also fixes plethora of bugs related to
subsequent PutObject() races with namespace locking.

This patch also enables most of the tests, other than
ListObjects paging which has some issues still.

Fixes #1358, #1360
2016-04-25 12:47:31 -07:00
Harshavardhana cab6805f09 xl: Enable a subset of tests for XL branch. (#1359) 2016-04-25 12:47:31 -07:00
Krishna Srinivas 8c85815106 xl: refactor functions to xl-v1-common.go xl-v1-utils.go. (#1357) 2016-04-25 12:47:31 -07:00
Krishna Srinivas becc814531 Xl layer selfheal quorum2
* xl/selfheal: selfheal based on read quorum on GET

* xl: getReadableDisks() also returns whether self-heal is needed so that this info can be used by ReadFile/SelfHeal/StatFile.

* xl: trigger selfheal from StatFile.
2016-04-25 12:47:31 -07:00
Harshavardhana 9bd9441107 xl: Simplify reading metadata and add a new fileMetadata type. (#1346) 2016-04-25 12:47:31 -07:00
Harshavardhana f3784d1087 xl: Handle read quorum for StatVol, ListVols 2016-04-25 12:47:31 -07:00
Harshavardhana 91588209fa obj: Object api handle all errors in common location. (#1343) 2016-04-25 12:47:31 -07:00
Krishna Srinivas 5c33b68318 xl: code refactor, cleanup ReadFile and CreateFile. 2016-04-25 12:47:31 -07:00
Bala FA 45b3d3e21f xl: add quorum support for create file 2016-04-25 12:47:31 -07:00
Harshavardhana 141a44bfbf xl: Fix ReadFile to keep the order always for reading the data back. (#1339)
Also fixes a stackoverflow bug in namespace locking.
2016-04-25 12:47:31 -07:00
Harshavardhana c7bf471c9e list/xl: Fix the way marker is handled in leafDirectory verification. 2016-04-25 12:47:31 -07:00
Krishna Srinivas c302875774 selfheal: implement self-heal. Heals the missing parts. (#1335) 2016-04-25 12:47:31 -07:00
Harshavardhana b76f3f1d62 xl: Add more fixes and cleanup.
Simplify cleanup of temporary files during createFile operations.
2016-04-25 12:47:31 -07:00
Bala FA ada0f82b9a xl: add quorum support for read file and name space locking. (#1333) 2016-04-25 12:47:31 -07:00
Harshavardhana a98a7fb1ad Implement XL layer - preliminary work. 2016-04-25 12:47:31 -07:00
Harshavardhana bf8a9702a4 tests: Fix a bug in TestObjectAPIIsUploadIDExists. (#1375)
The following code crashes when upload ID does not
exist, since we are setting err == nil when we find
err == errFileNotFound.

```
if e == nil {
   t.Fatal(e.Error())
```

Fix it.
2016-04-25 12:47:08 -07:00
karthic rao 6e372f83b4 Tests: object api multipart tests and bug fixes. 2016-04-25 10:39:28 -07:00
Harshavardhana e9fba04b36 logging: Enable logging across storage fs layer. (#1367)
Adds log.Debugf at all the layer - fixes #1074
2016-04-24 00:36:00 -07:00
Harshavardhana d63d17012d tests: Add API suite tests back for object api. (#1352) 2016-04-21 23:40:01 -07:00
Harshavardhana 444d1f8a65 miniobrowser: Vendorize to new changes in miniobrowser. 2016-04-21 20:35:48 -07:00
karthic rao 560c3bd153 Adding return statement after error response in the lastest commit to verify location constraint (#1348) 2016-04-21 20:08:08 -07:00
Harshavardhana 4cf73caf02 api: Add diskInfo as part of StatVol and ListVols. (#1349)
It is the bucket and volumes which needs to have this
value rather than the DiskInfo API itself. Eventually
this can be extended to show disk usage per
Buckets/Volumes whenever we have that functionality.

For now since buckets/volumes are thinly provisioned
this is the right approach.
2016-04-21 20:07:47 -07:00
Harshavardhana 1284ecc6f2 api: Fix verification of checkLeafDirectory. (#1347)
This fixes a problem where leaf directory has more than 1000
entries, also resulting in listing issues, leading to an infinite
loop.

Fixes #1334
2016-04-21 18:05:26 -07:00
karthic rao cb1116725b api: verify Location constraint for make bucket. (#1342) 2016-04-20 17:35:38 -07:00
koolhead17 c3d0a3d51e Update README.md
Our community contributor bought this to our attention so we have to add region as well in s3cmd config file.
2016-04-19 10:22:40 -07:00
Harshavardhana e0f8fed011 object: handle Error responses and handle errDiskFull. (#1331) 2016-04-19 02:42:10 -07:00
Harshavardhana 6bc17a3aea server: Attempt to increase max open files. (#1328) 2016-04-18 22:05:32 -07:00
Dave Henderson ff9a6b00cc isContainerized should look for /.dockerenv not /.dockerinit
Signed-off-by: Dave Henderson <dhenderson@gmail.com>
2016-04-18 19:26:56 -07:00
matt robinson af907a35a9 add environment var to explicitly indicate containerized and allow running as root (#1327) 2016-04-18 16:30:11 -07:00
Harshavardhana b47d722d8e fs: Fix filtering out valid paths from previous #1321 fix (#1323)
Fixes #1324
2016-04-17 12:00:23 -07:00
Harshavardhana 33633fd15d fs: Filter out valid paths and volnames. (#1321) 2016-04-16 19:33:29 -07:00
Harshavardhana b2ec7da9f9 Merge pull request #1314 from minio/split-fs
Split fs code into storage layer and object layer.
2016-04-16 18:09:31 -07:00
Harshavardhana be002ac01e fs/object: Fix issues from review comments. 2016-04-16 17:57:14 -07:00
Krishna Srinivas 149c6ca094 listMultipart: bugfixes. (#1318) 2016-04-16 16:25:53 -07:00
Harshavardhana 8457af5708 fs: Add proper volume and path validation. 2016-04-16 16:25:53 -07:00
Krishna Srinivas caa35f68fa listMultipart: implement support for marker. (#1313) 2016-04-16 16:25:53 -07:00
Harshavardhana 30b0b4deba storage/server/client: Enable storage server, enable client storage. 2016-04-16 16:25:53 -07:00
Krishna Srinivas 01a439f95b refactor: add multipart code to the object layer. 2016-04-16 16:25:53 -07:00
Krishna Srinivas 3c48537f20 refactor: refactor code to separate fs into object-layer and fs layer. (#1305) 2016-04-16 16:25:53 -07:00
karthic rao 188bb92d8a bucket-policy parset tests, and bug fixes (#1317) 2016-04-15 18:23:19 -07:00
Harshavardhana 6b3fc03701 docker: Update docker document 2016-04-15 17:40:56 -07:00
Harshavardhana 8112291d43 Add FreeBSD binary link and make a release 2016-04-14 17:22:47 -07:00
Harshavardhana 93666827f4 release: Add freebsd/amd64 build and remove zip, tgz. (#1316) 2016-04-13 23:34:55 -07:00
GarimaKapoor ac30bef72a Revised Docker.md (#1311) 2016-04-11 23:59:12 -07:00
Bala FA bea6f33b08 backend/fs: remove timer channel from scanMultipartDir() (#1310)
Previously scanMultipartDir() returns object info channel and timer
channel where timer channel is used to check whether object info
channel is alive or not.  This causes a race condition that timeout
may occur while object info channel in use.

This patch fixes the issue by removing timer channel and uses object
info channel directly where each object info has End bool field
indicates whether received object info is end or not.
2016-04-11 19:00:08 -07:00
Michael Werle 9fb1c79456 Improved Docker examples (#1308)
- Fixed a bug in the persistent docker command ("server" in place of "export")
- Added example of how to set consistent keys with ephemeral data, particularly useful for testing.
2016-04-11 12:33:34 -07:00
Harshavardhana 6b5699b15f config: console logging should be enabled by default. (#1307) 2016-04-09 14:19:59 -07:00