Commit Graph

134 Commits

Author SHA1 Message Date
Brendan Ashworth bccc386994 fs: drop Stat() call from fsDeleteFile,deleteFile (#4744)
This commit makes fsDeleteFile() simply call deleteFile() after calling
the relevant path length checking functions. This DRYs the code base.

This commit removes the Stat() call from deleteFile(). This improves
performance and removes any possibility of a race condition.

This additionally adds tests and a benchmark for said function. The
results aren't very consistent, although I'd expect this commit to make
it faster.
2017-08-03 20:04:28 -07:00
Harshavardhana cc8a8cb877 posix: Check for min disk space and inodes (#4618)
This is needed such that we don't start or
allow writing to a posix disk which doesn't
have minimum total disk space available.

One part fix for #4617
2017-07-10 18:14:48 -07:00
Harshavardhana 075b8903d7 fs: Add safe locking semantics for `format.json` (#4523)
This patch also reverts previous changes which were
merged for migration to the newer disk format. We will
be bringing these changes in subsequent releases. But
we wish to add protection in this release such that
future release migrations are protected.

Revert "fs: Migration should handle bucketConfigs as regular objects. (#4482)"
This reverts commit 976870a391.

Revert "fs: Migrate object metadata to objects directory. (#4195)"
This reverts commit 76f4f20609.
2017-06-12 17:40:28 -07:00
Frank Wessels 0f0758aece Load IO error count for posix atomically (#4448)
* Load error count atomically in order to check for maximum allowed number of IO errors.

* Remove unused (previously atomic) network IO error count
2017-05-31 09:22:53 -07:00
Aditya Manthramurthy 8975da4e84 Add new ReadFileWithVerify storage-layer API (#4349)
This is an enhancement to the XL/distributed-XL mode. FS mode is
unaffected.

The ReadFileWithVerify storage-layer call is similar to ReadFile with
the additional functionality of performing bit-rot checking. It
accepts additional parameters for a hashing algorithm to use and the
expected hex-encoded hash string.

This patch provides significant performance improvement because:

1. combines the step of reading the file (during
erasure-decoding/reconstruction) with bit-rot verification;

2. limits the number of file-reads; and

3. avoids transferring the file over the network for bit-rot
verification.

ReadFile API is implemented as ReadFileWithVerify with empty hashing
arguments.

Credits to AB and Harsha for the algorithmic improvement.

Fixes #4236.
2017-05-16 14:21:52 -07:00
Harshavardhana 76f4f20609 fs: Migrate object metadata to objects directory. (#4195)
Fixes #3352
2017-05-05 08:49:09 -07:00
Harshavardhana f0b5c0ec7c windows: Support all REPARSE_POINT attrib files properly. (#4203)
This change adopts the upstream fix in this regard at
https://go-review.googlesource.com/#/c/41834/ for Minio's
purposes.

Go's current os.Stat() lacks support for lot of strange
windows files such as

 - share symlinks on SMB2
 - symlinks on docker nanoserver
 - de-duplicated files on NTFS de-duplicated volume.

This PR attempts to incorporate the change mentioned here

   https://blogs.msdn.microsoft.com/oldnewthing/20100212-00/?p=14963/

The article suggests to use Windows I/O manager to
dereference the symbolic link.

Fixes #4122
2017-05-02 02:35:27 -07:00
Anis Elleuch 79e0b9e69a Relax minio server start when disk threshold is reached and adds space check in FS (#3865)
* fs: Rename tempObjPath variable in fsCreateFile()
* fs/posix: Factor checkDiskFree() function
* fs: Add disk free check in fsCreateFile()
* posix: Move free disk check to createFile()
* xl: Relax free disk check in POSIX initialization
* fs: checkDiskFree checks for space to store data
2017-03-07 12:25:40 -08:00
Harshavardhana 50b4e54a75 fs: Do not return reservedBucket names in ListBuckets() (#3754)
Make sure to skip reserved bucket names in `ListBuckets()`
current code didn't skip this properly and also generalize
this behavior for both XL and FS.
2017-02-16 14:52:14 -08:00
Krishnan Parthasarathi e5773e11c6 Make minio server compile on OpenBSD, NetBSD, Solaris (#3719) 2017-02-08 22:27:35 -08:00
Jeffery Utter 9e1f1b50e0 Don't Check Available Inodes on NFS (#3598)
In some cases (such as with VirutualBox, this value gets hardcoded
to 1000, which is less than the required minimum of 10000.

Fixes #3592
2017-01-19 10:39:44 -08:00
Harshavardhana 62f8343879 Add constants for commonly used values. (#3588)
This is a consolidation effort, avoiding usage
of naked strings in codebase. Whenever possible
use constants which can be repurposed elsewhere.

This also fixes `goconst ./...` reported issues.
2017-01-18 12:24:34 -08:00
Harshavardhana 1c699d8d3f fs: Re-implement object layer to remember the fd (#3509)
This patch re-writes FS backend to support shared backend sharing locks for safe concurrent access across multiple servers.
2017-01-16 17:05:00 -08:00
Harshavardhana 2062add05f fs/posix: On windows use helpers and init format.json properly. (#3434)
Fixes #3433
2016-12-12 15:43:41 -08:00
Krishnan Parthasarathi feb6685359 posix: Use preparePath only for paths used with syscall or os functions (#3377) 2016-11-30 20:56:15 -08:00
Harshavardhana 6efee2072d objectLayer: Check for `format.json` in a wrapped disk. (#3311)
This is needed to validate if the `format.json` indeed exists
when a fresh node is brought online.

This wrapped implementation also connects to the remote node
by attempting a re-login. Subsequently after a successful
connect `format.json` is validated as well.

Fixes #3207
2016-11-23 15:48:10 -08:00
Bala FA 825000bc34 Use humanize constants for KiB, MiB and GiB units. (#3322) 2016-11-22 18:18:22 -08:00
Anis Elleuch 4098025c11 Remove XL multipart tmp files when the latter is canceled (#3214)
XL multipart fails to remove tmp files when an error occurs during upload, this case covers the scenario where an upload is canceled manually by the client in the middle of job.
2016-11-21 16:34:57 -08:00
Krishna Srinivas afa4c7c3ef fs/multipart: Append multipart parts in a proper Go routine in background. (#3282) 2016-11-20 23:42:53 -08:00
Harshavardhana 0b9f0d14a1 auth/rpc: Take remote disk offline after maximum allowed attempts. (#3288)
Disks when are offline for a long period of time, we should
ignore the disk after trying Login upto 5 times.

This is to reduce the network chattiness, this also reduces
the overall time spent on `net.Dial`.

Fixes #3286
2016-11-20 16:57:12 -08:00
Harshavardhana 2f7fb78692 rpc: Our rpcClient should make an attempt to reconnect. (#3221)
rpcClient should attempt a reconnect if the call fails
with 'rpc.ErrShutdown' this is needed since at times when
the servers are taken down and brought back up.

The hijacked connection from net.Dial is usually closed.

So upon first attempt rpcClient might falsely indicate that
disk to be down, to avoid this state make another dial attempt
to really fail.

Fixes #3206
Fixes #3205
2016-11-10 07:44:41 -08:00
Harshavardhana f3c6c55719 posix: Fix windows performance issues. (#3132)
Do not attempt to fetch volume/drive information for
each i/o situation. In our case we do this in all calls
`posix.go` this in-turn created a terrible situation for
windows. This issue does not affect the i/o path on Unix
platforms since statvfs calls are in the range of micro
seconds on these platforms.

This verification is only needed during startup and we
let things fail at a later stage on windows.
2016-10-31 09:34:44 -07:00
Anis Elleuch a47ce7ab22 Add support of fallocate for FS and XL backends (#3032) 2016-10-29 12:44:44 -07:00
Harshavardhana 9e2d0ac50b Move to URL based syntax formatting. (#3092)
For command line arguments we are currently following

- <node-1>:/path ... <node-n>:/path

This patch changes this to

- http://<node-1>/path ... http://<node-n>/path
2016-10-27 03:30:52 -07:00
Harshavardhana e9c45102b0 posix: Use sync.Pool buffers to copy in large buffers. (#3106)
These fixes are borrowed from the fixes required for GlusterFS i/o throughput.
2016-10-26 17:14:05 -07:00
Krishna Srinivas d3aaf50a40 posix: Split on ":" in path d:\export makes minio use wrong disk. (#3027)
As the host/path split happens at a higher layer now, split at posix is not needed.
fixes part of #2987
2016-10-20 23:39:33 -07:00
Harshavardhana 95567c68bf posix: Do not print errors in expected errors. (#3012)
Fixes #3011
2016-10-20 09:26:18 -07:00
Harshavardhana 8d2347bc7b storage: DeleteFile should return errFileNotFound for ENOENT. (#2978) 2016-10-17 16:38:46 -07:00
Harshavardhana 6494b77d41 server: Add more elaborate startup messages. (#2731)
These messages based on our prep stage during XL
and prints more informative message regarding
drive information.

This change also does a much needed refactoring.
2016-10-05 12:48:07 -07:00
Anis Elleuch 9417614a8e Recalculate free minimum disk space (#2788)
* Fix calculating free space disk by using blocks available for unprivileged user

* Use fixed minimal free disk space instead of percentage
2016-09-27 12:46:38 -07:00
Anis Elleuch d936ed90ae Avoid testing on system errors strings in posix (#2583) 2016-09-13 21:18:30 -07:00
Harshavardhana 339425fd52 server: Fetch StorageInfo() from underlying disks transparently. (#2549)
Fixes #2511
2016-09-13 21:18:30 -07:00
Mohit Agarwal 418921de89 minor cleanup
- Reused contains() from utils.go at a couple of places
- Cleanup in return statements and boolean checks
2016-08-24 22:54:34 -07:00
Harshavardhana bccf549463 server: Move all the top level files into cmd folder. (#2490)
This change brings a change which was done for the 'mc'
package to allow for clean repo and have a cleaner
github drop in experience.
2016-08-18 16:23:42 -07:00