Commit Graph

194 Commits

Author SHA1 Message Date
Klaus Post
9004d69c6f
Make ReqInfo concurrency safe (#15204)
Some read/writes of ReqInfo did not get appropriate locks, leading to races.

Make sure reading and writing holds appropriate locks.
2022-06-30 10:48:50 -07:00
Harshavardhana
9c605ad153
allow support for parity '0', '1' enabling support for 2,3 drive setups (#15171)
allows for further granular setups

- 2 drives (1 parity, 1 data)
- 3 drives (1 parity, 2 data)

Bonus: allows '0' parity as well.
2022-06-27 20:22:18 -07:00
Harshavardhana
bd099f5e71
fix: change timedValue to return the previously cached value (#15169)
fix: change timedvalue to return previous cached value

caller can interpret the underlying error and decide
accordingly, places where we do not interpret the
errors upon timedValue.Get() - we should simply use
the previously cached value instead of returning "empty".

Bonus: remove some unused code
2022-06-25 08:50:16 -07:00
Harshavardhana
1a40c7c27c
use signature-v2 for 'object perf' tests to avoid CPU using sha256 (#15151)
It is observed in a local 8 drive system the CPU seems to be
bottlenecked at

```
(pprof) top
Showing nodes accounting for 1385.31s, 88.47% of 1565.88s total
Dropped 1304 nodes (cum <= 7.83s)
Showing top 10 nodes out of 159
      flat  flat%   sum%        cum   cum%
      724s 46.24% 46.24%       724s 46.24%  crypto/sha256.block
   219.04s 13.99% 60.22%    226.63s 14.47%  syscall.Syscall
   158.04s 10.09% 70.32%    158.04s 10.09%  runtime.memmove
   127.58s  8.15% 78.46%    127.58s  8.15%  crypto/md5.block
    58.67s  3.75% 82.21%     58.67s  3.75%  github.com/minio/highwayhash.updateAVX2
    40.07s  2.56% 84.77%     40.07s  2.56%  runtime.epollwait
    33.76s  2.16% 86.93%     33.76s  2.16%  github.com/klauspost/reedsolomon._galMulAVX512Parallel84
     8.88s  0.57% 87.49%     11.56s  0.74%  runtime.step
     7.84s   0.5% 87.99%      7.84s   0.5%  runtime.memclrNoHeapPointers
     7.43s  0.47% 88.47%     22.18s  1.42%  runtime.pcvalue
```

Bonus changes:

- re-use transport for bucket replication clients, also site replication clients.
- use 32KiB buffer for all read and writes at transport layer seems to help
  TLS read connections.
- Do not have 'MaxConnsPerHost' this is problematic to be used with net/http
  connection pooling 'MaxIdleConnsPerHost' is enough.
2022-06-22 16:28:25 -07:00
Andreas Auernhammer
cd7a0a9757
fips: simplify TLS configuration (#15127)
This commit simplifies the TLS configuration.
It inlines the FIPS / non-FIPS code.

Signed-off-by: Andreas Auernhammer <hi@aead.dev>
2022-06-21 07:54:48 -07:00
Harshavardhana
f1abb92f0c
feat: Single drive XL implementation (#14970)
Main motivation is move towards a common backend format
for all different types of modes in MinIO, allowing for
a simpler code and predictable behavior across all features.

This PR also brings features such as versioning, replication,
transitioning to single drive setups.
2022-05-30 10:58:37 -07:00
Harshavardhana
c7df1ffc6f
avoid concurrent reads and writes to opts.UserDefined (#14862)
do not modify opts.UserDefined after object-handler
has set all the necessary values, any mutation needed
should be done on a copy of this value not directly.

As there are other pieces of code that access opts.UserDefined
concurrently this becomes challenging.

fixes #14856
2022-05-05 04:14:41 -07:00
Anis Elleuch
44a3b58e52
Add audit log for decommissioning (#14858) 2022-05-04 00:45:27 -07:00
Harshavardhana
16bc11e72e
fix: disallow newer policies, users & groups with space characters (#14845)
space characters at the beginning or at the end can lead to
confusion under various UI elements in differentiating the
actual name of "policy, user or group" - to avoid this behavior
this PR onwards we shall reject such inputs for newer entries.

existing saved entries will behave as is and are going to be
operable until they are removed/renamed to something more
meaningful.
2022-05-02 09:27:35 -07:00
Klaus Post
13ba4b433d
Clean up cpuio profiling (#14838)
Don't start regular cpu profile as well. Use bed madmin const.
2022-04-29 09:35:42 -07:00
Aditya Manthramurthy
0e502899a8
Add support for multiple OpenID providers with role policies (#14223)
- When using multiple providers, claim-based providers are not allowed. All
providers must use role policies.

- Update markdown config to allow `details` HTML element
2022-04-28 18:27:09 -07:00
Harshavardhana
eda34423d7 update gofumpt -w - new changes 2022-04-13 12:00:11 -07:00
Harshavardhana
92a77cc78e
update pkg v1.1.20 to reload certs in k8s always (#14470) 2022-03-04 20:34:39 -08:00
Klaus Post
b88c57ba93
Add fgprof profiles (#14321)
https://github.com/felixge/fgprof#rocket-fgprof---the-full-go-profiler
2022-02-16 12:00:10 -08:00
Sidhartha Mani
d7df6bc738
add support for speedtest drive (#14182) 2022-02-01 22:38:05 -08:00
Harshavardhana
57118919d2
cached diskIDs are not needed for scanner healing (#14170)
This PR removes an unnecessary state that gets
passed around for DiskIDs, which is not necessary
since each disk exactly knows which pool and which
set it belongs to on a running system.

Currently cached DiskId's won't work properly
because it always ends up skipping offline disks
and never runs healing when disks are offline, as
it expects all the cached diskIDs to be present
always. This also sort of made things in-flexible
in terms perhaps a new diskID for `format.json`.
(however this is not a big issue)

This is an unnecessary requirement that healing
via scanner needs all drives to be online, instead
healing should trigger even when partial nodes
and drives are available this ensures that we
keep the SLA in-tact on the objects when disks
are offline for a prolonged period of time.
2022-01-26 08:34:56 -08:00
Aditya Manthramurthy
0a224654c2
fix: progagation of service accounts for site replication (#14054)
- Only non-root-owned service accounts are replicated for now.
- Add integration tests for OIDC with site replication
2022-01-07 17:41:43 -08:00
Minio Trusted
76877eb6fa move gofumpt to golang-ci 2022-01-06 13:08:21 -08:00
Klaus Post
3d66d053c7
Add small client TLS PSK cache (#14039) 2022-01-06 11:34:02 -08:00
Harshavardhana
a60ac7ca17
fix: audit log to support object names in multipleObjectNames() handler (#14017) 2022-01-03 01:28:52 -08:00
Harshavardhana
b9aae1aaae
fix: speedtest should exit upon errors cleanly (#13851)
- deleteBucket() should be called for cleanup
  if client abruptly disconnects

- out of disk errors should be sent to client
  properly and also cancel the calls

- limit concurrency to available MAXPROCS not
  32 for auto-tuned setup, if procs are beyond
  32 then continue normally. this is to handle
  smaller setups.

fixes #13834
2021-12-06 16:36:14 -08:00
Harshavardhana
24d904d194
reload certs from disk upon SIGHUP (#13792) 2021-12-01 00:38:32 -08:00
Harshavardhana
e49c184595
add configurable 'shutdown-timeout' for HTTP server (#13771)
fixes #12317
2021-11-29 09:06:56 -08:00
Harshavardhana
99d87c5ca2
fix: totalDrives reported in speedTest for multiple-pools (#13770)
totalDrives reported in speedTest result were wrong
for multiple pools, this PR fixes this.

Bonus: add support for configurable storage-class, this
allows us to test REDUCED_REDUNDANCY to see further
maximum throughputs across the cluster.
2021-11-29 09:05:46 -08:00
Harshavardhana
661b263e77
add gocritic/ruleguard checks back again, cleanup code. (#13665)
- remove some duplicated code
- reported a bug, separately fixed in #13664
- using strings.ReplaceAll() when needed
- using filepath.ToSlash() use when needed
- remove all non-Go style comments from the codebase

Co-authored-by: Aditya Manthramurthy <donatello@users.noreply.github.com>
2021-11-16 09:28:29 -08:00
Harshavardhana
c3d24fb26d
use single encoder for sending speedtest results (#13579)
Bonus: if runs have PUT higher then capture it anyways
to display an unexpected result, which provides a way
to understand what might be slowing things down on the
system.

For example on a Data24 WDC setup it is clearly visible
there is a bug in the hardware.

```
./mc admin speedtest wdc/
⠧ Running speedtest (With 64 MiB object size, 32 concurrency) PUT: 31 GiB/s GET: 24 GiB/s
⠹ Running speedtest (With 64 MiB object size, 48 concurrency) PUT: 38 GiB/s GET: 24 GiB/s

MinIO 2021-11-04T06:08:33Z, 6 servers, 48 drives
PUT: 38 GiB/s, 605 objs/s
GET: 24 GiB/s, 383 objs/s
```

Reads are almost 14GiB/sec slower than Writes which
is practically not possible.
2021-11-04 12:11:52 -07:00
Krishna Srinivas
58934e5881
Support live updates for clients during speedtest (#13566) 2021-11-02 15:27:03 -07:00
Krishna Srinivas
70160aeab3
Remove IOPS autotuning and simplify autotune code (#13554) 2021-11-02 13:03:00 -07:00
Klaus Post
974073a2e5
directio: Check if buffers are set. (#13440)
Check if directio buffers have actually been fetched and prevent errors on double Close. Return error on Read after Close.

Fixes

```
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0xf8582f]

goroutine 210 [running]:
github.com/minio/minio/internal/ioutil.(*ODirectReader).Read(0xc0054f8320, {0xc0014560b0, 0xa8, 0x44d012})
	github.com/minio/minio/internal/ioutil/odirect_reader.go:88 +0x10f
io.ReadAtLeast({0x428c5c0, 0xc0054f8320}, {0xc0014560b0, 0xa8, 0xa8}, 0xa8)
	io/io.go:328 +0x9a
io.ReadFull(...)
	io/io.go:347
github.com/minio/minio/internal/ioutil.ReadFile({0xc001bf60e0, 0x6})
	github.com/minio/minio/internal/ioutil/read_file.go:48 +0x19b
github.com/minio/minio/cmd.(*FSObjects).scanBucket.func1({{0xc00444e1e0, 0x4d}, 0x0, {0xc0040cf240, 0xe}, {0xc0040cf24f, 0x18}, {0xc0040cf268, 0x18}, 0x0, ...})
	github.com/minio/minio/cmd/fs-v1.go:366 +0x1ea
github.com/minio/minio/cmd.(*folderScanner).scanFolder.func1({0xc00474a6a8, 0xc0065d6793}, 0x0)
	github.com/minio/minio/cmd/data-scanner.go:494 +0xb15
github.com/minio/minio/cmd.readDirFn({0xc002803e80, 0x34}, 0xc000670270)
	github.com/minio/minio/cmd/os-readdir_unix.go:172 +0x638
github.com/minio/minio/cmd.(*folderScanner).scanFolder(0xc002deeb40, {0x42dc9d0, 0xc00068cbc0}, {{0xc001c6e2d0, 0x27}, 0xc0023db8e0, 0x1}, 0xc0001c7ab0)
	github.com/minio/minio/cmd/data-scanner.go:427 +0xa8f
github.com/minio/minio/cmd.(*folderScanner).scanFolder.func2({{0xc001c6e2d0, 0x27}, 0xc0023db8e0, 0x27})
	github.com/minio/minio/cmd/data-scanner.go:549 +0xd0
github.com/minio/minio/cmd.(*folderScanner).scanFolder(0xc002deeb40, {0x42dc9d0, 0xc00068cbc0}, {{0xc0013fa9e0, 0xe}, 0x0, 0x1}, 0xc000670dd8)
	github.com/minio/minio/cmd/data-scanner.go:623 +0x205d
github.com/minio/minio/cmd.scanDataFolder({_, _}, {_, _}, {{{0xc0013fa9e0, 0xe}, 0x802, {0x210f15d2, 0xed8f903b8, 0x5bc0e80}, ...}, ...}, ...)
	github.com/minio/minio/cmd/data-scanner.go:333 +0xc51
github.com/minio/minio/cmd.(*FSObjects).scanBucket(_, {_, _}, {_, _}, {{{0xc0013fa9e0, 0xe}, 0x802, {0x210f15d2, 0xed8f903b8, ...}, ...}, ...})
	github.com/minio/minio/cmd/fs-v1.go:364 +0x305
github.com/minio/minio/cmd.(*FSObjects).NSScanner(0x42dc9d0, {0x42dc9d0, 0xc00068cbc0}, 0x0, 0xc003bcfda0, 0x802)
	github.com/minio/minio/cmd/fs-v1.go:307 +0xa16
github.com/minio/minio/cmd.runDataScanner({0x42dc9d0, 0xc00068cbc0}, {0x436a6c0, 0xc000bfcf50})
	github.com/minio/minio/cmd/data-scanner.go:150 +0x749
created by github.com/minio/minio/cmd.initDataScanner
	github.com/minio/minio/cmd/data-scanner.go:73 +0xb0
```
2021-10-14 10:19:17 -07:00
Krishna Srinivas
03a2a74697
Support speedtest autotune on the server side (#13086) 2021-09-10 17:43:34 -07:00
Harshavardhana
e124d88788
optimize listing operation concurrency (#12728)
- remove use of getOnlineDisks() instead rely on fallbackDisks()
  when disk return errors like diskNotFound, unformattedDisk
  use other fallback disks to list from, instead of paying the
  price for checking getOnlineDisks()

- optimize getDiskID() further to avoid large write locks when
  looking formatLastCheck time window

This new change allows for a more relaxed fallback for listing
allowing for more tolerance and also eventually gain more
consistency in results even if using '3' disks by default.
2021-07-24 22:03:38 -07:00
Anis Elleuch
23ef25b57a
profiling: Return goroutines with sleep duration (#12775)
Add a new goroutine file which has another printing format. We need it
to see how much time each goroutine was blocked. Easier to detect stops.

Co-authored-by: Anis Elleuch <anis@min.io>
2021-07-23 13:16:53 -07:00
Poorna Krishnamoorthy
a69c2a2fb3
Change replication to use read lock instead of writelock (#12581)
Fixes #12573

This PR also adding audit logging for replication activity
2021-06-28 23:58:08 -07:00
Poorna Krishnamoorthy
d00783c923
Use rate.Limiter for bandwidth monitoring (#12506)
Bonus: fixes a hang when bandwidth caps are enabled for
synchronous replication
2021-06-24 18:29:30 -07:00
Harshavardhana
cdeccb5510
feat: Deprecate embedded browser and import console (#12460)
This feature also changes the default port where
the browser is running, now the port has moved
to 9001 and it can be configured with

```
--console-address ":9001"
```
2021-06-17 20:27:04 -07:00
Harshavardhana
1f262daf6f
rename all remaining packages to internal/ (#12418)
This is to ensure that there are no projects
that try to import `minio/minio/pkg` into
their own repo. Any such common packages should
go to `https://github.com/minio/pkg`
2021-06-01 14:59:40 -07:00
Harshavardhana
81d5688d56
move the dependency to minio/pkg for common libraries (#12397) 2021-05-28 15:17:01 -07:00
Andreas Auernhammer
d8eb7d3e15
kms: replace KES client implementation with minio/kes (#12207)
This commit replaces the custom KES client implementation
with the KES SDK from https://github.com/minio/kes

The SDK supports multi-server client load-balancing and
requests retry out of the box. Therefore, this change reduces
the overall complexity within the MinIO server and there
is no need to maintain two separate client implementations.

Signed-off-by: Andreas Auernhammer <aead@mail.de>
2021-05-10 18:15:11 -07:00
Harshavardhana
1aa5858543
move madmin to github.com/minio/madmin-go (#12239) 2021-05-06 08:52:02 -07:00
Harshavardhana
069432566f update license change for MinIO
Signed-off-by: Harshavardhana <harsha@minio.io>
2021-04-23 11:58:53 -07:00
Harshavardhana
bb1198c2c6
revert CreateFile waitForResponse (#12124)
instead use expect continue timeout, and have
higher response header timeout, the new higher
timeout satisfies worse case scenarios for total
response time on a CreateFile operation.

Also set the "expect" continue header to satisfy
expect continue timeout behavior.

Some clients seem to cause CreateFile body to be
truncated, leading to no errors which instead
fails with ObjectNotFound on a PUT operation,
this change avoids such failures appropriately.

Signed-off-by: Harshavardhana <harsha@minio.io>
2021-04-23 10:18:18 -07:00
Poorna Krishnamoorthy
c9bf6007b4
Use custom transport for remote targets (#12080) 2021-04-16 18:58:26 -07:00
Harshavardhana
a334554f99
fix: add helper for expected path.Clean behavior (#12068)
current usage of path.Clean returns "." for empty strings
instead we need `""` string as-is, make relevant changes
as needed.
2021-04-15 16:32:13 -07:00
Harshavardhana
feafccf007
handle trimming '/' if present in the object names (#11765)
- MultipleDeletes should handle '/' prefix for objectnames
- Trimming the slash alone is enough for ListObjects()
  prefix and markers

fixes #11769
2021-03-11 13:57:03 -08:00
Harshavardhana
691035832a
fix: normalize object layer inputs (#11534)
Cases where we have applications making request
for `//` in object names make sure that all
are normalized to `/` and all such requests that
are prefixed '/' are removed. To ensure a
consistent view from all operations.
2021-03-09 12:58:22 -08:00
Harshavardhana
9ccc483df6
[feat]: change erasure coding default block size from 10MiB to 1MiB (#11721)
major performance improvements in range GETs to avoid large
read amplification when ranges are tiny and random

```
-------------------
Operation: GET
Operations: 142014 -> 339421
Duration: 4m50s -> 4m56s
* Average: +139.41% (+1177.3 MiB/s) throughput, +139.11% (+658.4) obj/s
* Fastest: +125.24% (+1207.4 MiB/s) throughput, +132.32% (+612.9) obj/s
* 50% Median: +139.06% (+1175.7 MiB/s) throughput, +133.46% (+660.9) obj/s
* Slowest: +203.40% (+1267.9 MiB/s) throughput, +198.59% (+753.5) obj/s
```

TTFB from 10MiB BlockSize
```
* First Access TTFB: Avg: 81ms, Median: 61ms, Best: 20ms, Worst: 2.056s
```

TTFB from 1MiB BlockSize
```
* First Access TTFB: Avg: 22ms, Median: 21ms, Best: 8ms, Worst: 91ms
```

Full object reads however do see a slight change which won't be
noticeable in real world, so not doing any comparisons

TTFB still had improvements with full object reads with 1MiB

```
* First Access TTFB: Avg: 68ms, Median: 35ms, Best: 11ms, Worst: 1.16s
```

v/s

TTFB with 10MiB
```
* First Access TTFB: Avg: 388ms, Median: 98ms, Best: 20ms, Worst: 4.156s
```

This change should affect all new uploads, previous uploads should
continue to work with business as usual. But dramatic improvements can
be seen with these changes.
2021-03-06 14:09:34 -08:00
Andreas Auernhammer
f14cc6c943
etag: add FromContentMD5 to parse content-md5 as ETag (#11688)
This commit adds the `FromContentMD5` function to
parse a client-provided content-md5 as ETag.

Further, it also adds multipart ETag computation
for future needs.
2021-03-03 12:58:28 -08:00
Anis Elleuch
e8d8dfa3ae
Add metric for internode RPC calls errors (#11669) 2021-03-01 12:31:33 -08:00
Nitish Tiwari
bbd1244a88
Add support for mTLS for Audit log target (#11645) 2021-03-01 09:19:13 -08:00
Klaus Post
c5b3a675fa
Block profiling tweaks (#11612)
The base profiles contains no valuable data, don't record them.

Reduce block rate by 2 orders of magnitude, should still capture just as valuable data with less CPU strain.
2021-02-27 09:22:14 -08:00