Commit Graph
849 Commits
Author SHA1 Message Date
Karthic RaoandHarshavardhana 8bffa78f7f Fix Instance type during benchmarks. (#3147)
- The benchmark initialization function was not taking into account the
  instance type (FS/XL), was using XL ObjectLayer even for FS
  benchmarks.
- This was leading to incorrect benchmark results for FS related
  benchmarks.
- The fix takes into account the instance type (FS/XL) and correctly
  returns FS backend for FS benchmarks.
2016-11-01 10:21:16 -07:00
Anis ElleuchandHarshavardhana 4b302173ae fs: Remove object append file if found since it is not always used (#3139) 2016-11-01 00:56:03 -07:00
Karthic RaoandHarshavardhana 9db2205db3 Fix List-Object Benchmark failure. (#3142) 2016-10-31 13:57:30 -07:00
Aditya ManthramurthyandHarshavardhana 6914fe436c Remove unused function. (#3143) 2016-10-31 12:18:15 -07:00
Anis ElleuchandHarshavardhana 807cc3c28d Tracing locking errors for better debugging (#3140) 2016-10-31 10:17:14 -07:00
HarshavardhanaandGitHub 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
Karthic RaoandHarshavardhana 3e8cb8c937 build: ineffassign fixes. (#3134) 2016-10-30 23:32:46 -07:00
Aditya ManthramurthyandHarshavardhana dd6ecf1193 Read/write quorum algo with uploads.json in xl (Fixes #3123) (#3124)
- Reads and writes of uploads.json in XL now uses quorum for
  newMultipart, completeMultipart and abortMultipart operations.

- Each disk's `uploads.json` file is read and updated independently for
  adding or removing an upload id from the file. Quorum is used to
  decide if the high-level operation actually succeeded.

- Refactor FS code to simplify the flow, and fix a bug while reading
  uploads.json.
2016-10-30 09:27:29 -07:00
Anis ElleuchandHarshavardhana a47ce7ab22 Add support of fallocate for FS and XL backends (#3032) 2016-10-29 12:44:44 -07:00
Krishna SrinivasandHarshavardhana 0b3282ac9f Logging: errorIf fatalIf print in the format [file.go:82:funcName()] (#3127) 2016-10-29 02:34:16 -07:00
Krishna SrinivasandHarshavardhana 79b98b5c25 boot: getPath() should take care of simple directory exports. (#3122)
fixes #3120
2016-10-28 13:08:56 -07:00
Krishnan ParthasarathiandHarshavardhana 6a57f2c1f0 XL: Add more information to panic msg (#3119) 2016-10-28 08:46:03 -07:00
Krishna SrinivasandHarshavardhana 0fa2477cb0 boot: check parameter syntax before initializing the system. (#3114)
fixes #3112
2016-10-28 08:45:32 -07:00
Anis ElleuchandHarshavardhana 7ffb337cd7 Update control cmd USAGE to indicate that it is possible to put access secret keys in URL (#3115) 2016-10-27 13:26:11 -07:00
HarshavardhanaandGitHub 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
Aditya ManthramurthyandHarshavardhana 30dc11a931 No listener.json for single-node mode (Fixes #3052) (#3108)
In FS or single-node XL mode, there is no need to save listener
configuration to persistent storage. As there is only one server, if it
is restarted, any connected listenBucketAPI clients were disconnected
and will have to reconnect - so there is nothing to actually store.

This incidentally solves #3052 by avoiding the problem.
2016-10-26 20:13:00 -07:00
Anis ElleuchandHarshavardhana a15dc5fed5 Print message when creating the config file (#3089) 2016-10-26 18:44:22 -07:00
Anis ElleuchandHarshavardhana f7c20b97a1 control cmds: Extract access and secret keys from URL if specified (#3109) 2016-10-26 18:42:12 -07:00
HarshavardhanaandGitHub 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
Anis ElleuchandHarshavardhana 8871eb8e1e Show offline nodes after a fixed number of init retry (#3107) 2016-10-26 16:09:06 -07:00
Krishna SrinivasandHarshavardhana 0f32efb825 PostPolicy - rename of files/functions + add testcases (#3104) 2016-10-26 10:15:57 -07:00
Karthic RaoandHarshavardhana 63f1b4fdf4 mispell fixes. (#3100) 2016-10-26 08:46:14 -07:00
Krishnan ParthasarathiandHarshavardhana 31f2db6880 Remove leftover debug statement from PutObject StreamingSignature unit-test (#3099) 2016-10-26 03:17:47 -07:00
Krishnan ParthasarathiandHarshavardhana 2c9b406f6c Add TLS based tests to functional test suite (#3083) 2016-10-26 02:30:31 -07:00
Krishnan ParthasarathiandHarshavardhana 12cd2da265 Add PutObjectHandler unit tests covering failure cases (#3096) 2016-10-26 02:06:22 -07:00
HarshavardhanaandGitHub 485c0ea8bf tests: Combine v2 tests with the Suite itself. (#3088) 2016-10-25 13:34:14 -07:00
Krishnan ParthasarathiandHarshavardhana 49ba07d1d6 Use net.ParseCIDR instead of custom-built parsers (#3055)
Removes avoidable conversion to and from net.IP to string.
2016-10-25 11:14:47 -07:00
Krishna SrinivasandHarshavardhana 35e541e0b1 content-length-range policy should be honored for the uploaded object sizes. (#3076) 2016-10-24 23:47:03 -07:00
Aditya ManthramurthyandHarshavardhana 3977d6b7bd Lock bucket while modifying its metadata (Fixes #2979) (#3019)
- When modifying notification configuration
- When modifying listener configuration
- When modifying policy configuration

With this change we also stop early checking if the bucket exists, since
that uses a Read-lock and causes a deadlock due to the outer Write-lock.
2016-10-24 19:52:24 -07:00
HarshavardhanaandGitHub 0905398459 Fix benchmark tests. (#3082)
Fixes #3081
2016-10-24 18:45:06 -07:00
Aditya ManthramurthyandHarshavardhana f41faf96b7 Fix newMultipartUpload to not leave stale uploads.json (Fixes #3014) (#3079) 2016-10-24 17:37:18 -07:00
Bala FAandHarshavardhana 36639b65a9 rename completeMultipartMD5() into getCompleteMultipartMD5(). (#3051) 2016-10-24 13:56:13 -07:00
HarshavardhanaandGitHub 7fc598b73f Fix user-agent prefix to have docker instead of suffix. (#3074) 2016-10-24 13:44:15 -07:00
Krishna SrinivasandHarshavardhana 21d41ad7fd init[windows]: Fix to handle the case when export path is a relative path. (#3054)
ex. to handle "minio server export"
2016-10-24 08:26:28 -07:00
HarshavardhanaandGitHub fe56220d1a Do not print nil when hostname is provided as --address (#3053)
Fixes #3018
2016-10-23 23:55:12 -07:00
HarshavardhanaandGitHub 5782ec3ada Fix peers and web UIVersion validation. (#3048) 2016-10-23 12:32:35 -07:00
Krishnan ParthasarathiandHarshavardhana 8839c5105a Pass values to closures esp. when passed to defer statement. (#3050)
opsID, a variable on the stack, changes over the course of
Completemultipartupload function in xl-v1-multipart.go.  This was
being used in a function closure which was passed to defer
statement. The variables used in the closure depend on their values at
the time of evaluation which is indeterminate behaviour. It is
incorrect to depend on values of variables on stack at the end of
function, when deferred functions are executed.
2016-10-23 09:57:52 -07:00
Anis ElleuchandHarshavardhana 6c2d5e3d05 Correct the number of failed disks that we can withstand in startup message (#3045) 2016-10-22 10:36:50 -07:00
Krishna SrinivasandHarshavardhana 5999a23d3e When object whose size is greater than 5G is uploaded using presigned POST we should return error. (#3033)
fixes #2961
2016-10-22 09:05:01 -07:00
Krishna SrinivasandHarshavardhana e51be73ac7 PresignedPost: Support for Signature V2 presigned POST Policy. (#3043)
fixes #2993
2016-10-22 08:57:12 -07:00
HarshavardhanaandGitHub 4b5b363c6c tests: Fix race between SetBucketListenerConfig and eventNotifyForBucketNotifications (#3041) 2016-10-22 02:35:33 -07:00
Krishna SrinivasandHarshavardhana f2b0c08e34 logging: print file:line:funcName of the caller of errorIf and fatalIf (#3035) 2016-10-22 02:24:34 -07:00
HarshavardhanaandGitHub 83b364891d tests: Fix a potential race in ListenBucketNotificationHandler. (#3040) 2016-10-21 22:56:27 -07:00
HarshavardhanaandGitHub e3ab478c70 tests: Fix a potential race in RemoveBucketNotification. (#3037)
Fixes #3036
2016-10-21 17:12:56 -07:00
HarshavardhanaandGitHub ecaccefd2e tests: Implement GetBucketNotification handler tests. (#3029) 2016-10-21 02:39:37 -07:00
HarshavardhanaandGitHub ece559afe2 api: Do not use sqs for ListenBucketNotification. (#3023)
Add more tests. Fixes #3024
2016-10-21 01:25:17 -07:00
Krishna SrinivasandHarshavardhana 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
Krishna SrinivasandHarshavardhana 32c3a558e9 distributed-XL: Support to run one minio process per export even on the same machine. (#2999)
fixes #2983
2016-10-20 18:31:02 -07:00
Anis ElleuchandHarshavardhana 41f9ab1c69 Translate storage access denied error to S3 Access Denied response (#3015) 2016-10-20 16:09:55 -07:00
Aditya ManthramurthyandHarshavardhana 8876e0a80a Delete bucket listener config file from disk (#3016) 2016-10-20 16:09:19 -07:00