github/ocf
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1,794 Commits 1 Branch 1 Tag
9c65ec955f43fbf386a3f3346e8e216ad2915798
Commit Graph

1794 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Michal Mielewczyk
9c65ec955f engine_rd: Ignore backfill buffer allocation error 
It's OK to proceed with a read even if failed to allocate a buffer for backfill

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-18 19:43:56 +02:00
Michal Mielewczyk
a3bccbba6c engine_rd: Refactor 
Code beautification only, no functional changes.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-18 19:38:51 +02:00
Robert Baldyga
5531b9cc79 Merge pull request #826 from mmichal10/errors-in-engines 
Errors in engines
 2024-09-18 15:26:21 +02:00
Rafal Stefanowski
194e5a9172 Use cache_error and core_error flags only in WT 
Signed-off-by: Rafal Stefanowski <rafal.stefanowski@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
 2024-09-18 14:04:08 +02:00
Rafal Stefanowski
2761540326 Report cache and core errors separately 
Signed-off-by: Rafal Stefanowski <rafal.stefanowski@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-18 08:01:08 +02:00
Robert Baldyga
b3f5ca1c2f Merge pull request #816 from mmichal10/zero-offset 
Support set_data() with offset > 0
 2024-09-17 21:00:35 +02:00
Roel Apfelbaum
73387c8f26 Support set_data() with offset > 0 for core 
Signed-off-by: Roel Apfelbaum <roel.apfelbaum@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-17 16:26:27 +02:00
Robert Baldyga
eb44557aed Merge pull request #824 from mmichal10/partial-hit-tests 
Partial hit tests
 2024-09-17 15:55:12 +02:00
Sara Merzel
9a6fe8a49b pyocf: Tests for partial hits 
Signed-off-by: Sara Merzel <sara.merzel@huawei.com>
Signed-off-by: Roel Apfelbaum <roel.apfelbaum@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-17 14:16:23 +02:00
Michal Mielewczyk
97477d484f pyocf: Helpers for sync IOs for Volume 
Signed-off-by: Sara Merzel <sara.merzel@huawei.com>
Signed-off-by: Roel Apfelbaum <roel.apfelbaum@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-17 10:04:02 +02:00
Michal Mielewczyk
f75ad04ec7 pyocf: Remove redundant imports 
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-12 14:07:30 +02:00
Robert Baldyga
d2fcd27d7a Merge pull request #823 from mmichal10/fix-ram-check 
Revert "Unstubify RAM check in posix evn"
 2024-09-12 12:41:30 +02:00
Michal Mielewczyk
60271ebb72 Revert "Unstubify RAM check in posix evn" 
This reverts commit 77d949bdcc.

Returning the actual amount of RAM may cause test_start_cache_huge_device to
fail

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-12 12:18:04 +02:00
Robert Baldyga
297f1cb8b0 Merge pull request #822 from mmichal10/fastpath_refactor 
Refactor resolving fast path
 2024-09-12 11:57:11 +02:00
Avi Halaf
bd06b1c9b8 Refactor resolving fast path 
Signed-off-by: Avi Halaf <avi.halaf@huawei.com>
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 16:47:40 +02:00
Robert Baldyga
d24e7bac5d Merge pull request #820 from mmichal10/discard-fixes 
Discard fixes
 2024-09-10 15:31:53 +02:00
Michal Mielewczyk
ca7f3651e9 discard engine: lookup without updating hotness 
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 15:20:51 +02:00
Michal Mielewczyk
0df0eec7f0 Uncouple lookup() and set_hot() 
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 15:20:51 +02:00
Rafal Stefanowski
7dfe70f69b Fix discard step callback refcount 
Signed-off-by: Rafal Stefanowski <rafal.stefanowski@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 15:20:51 +02:00
Robert Baldyga
070b16b31f Merge pull request #819 from mmichal10/rename-ops 
Rename engine_ops to engine_flush
 2024-09-10 15:19:51 +02:00
Robert Baldyga
1bcd949a89 Rename engine_ops to engine_flush 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 15:16:33 +02:00
Robert Baldyga
fa8c28ca10 Merge pull request #818 from mmichal10/dont-flush-volatile-md 
cleaner: skip metadata flush in volatile mode
 2024-09-10 12:44:38 +02:00
Amir Haroush
7930ef9c21 cleaner: skip metadata flush in volatile mode 
Signed-off-by: Amir Haroush <amir.haroush@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 12:32:25 +02:00
Robert Baldyga
bf679ecb49 Merge pull request #817 from mmichal10/env_ram_check 
Unstubify RAM check in posix evn
 2024-09-10 12:27:51 +02:00
Michal Mielewczyk
77d949bdcc Unstubify RAM check in posix evn 
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-10 10:24:15 +02:00
Robert Baldyga
ebbf832cdc Merge pull request #814 from mmichal10/debug-mechanism 
Debug mechanism
 2024-09-09 15:48:45 +02:00
Robert Baldyga
c07e44f036 pyocf: Fix cache settle 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
 2024-09-09 15:28:11 +02:00
Robert Baldyga
3ebdf38aa9 Introduce ocf_dbg_cache_is_settled() 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
 2024-09-09 15:28:11 +02:00
Michal Mielewczyk
f4d9f0dcf6 Introduce ocf_refcnt_zeroed() 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-09 15:28:09 +02:00
Robert Baldyga
1fbb00de8f Merge pull request #811 from mmichal10/req-rewrq 
Request improvements
 2024-09-09 14:37:03 +02:00
Robert Baldyga
dc58eeae9b Introduce d2c request 
This avoids unnecessary map allocation and initialization of unused fields of
request structure. It also allows to track thier number separately from
the regular requests

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-09 12:45:51 +02:00
Robert Baldyga
8b93b699c3 Eliminate queue -> cache mapping 
Eliminate need to resolve cache based on the queue. This allows to share
the queue between cache instances. The queue still holds pointer to
a cache that owns the queue, but no management or io path relies on the
queue -> cache mapping.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-09 12:45:51 +02:00
Robert Baldyga
460cd461d3 Allocate requests for management path separately 
Management path does not benefit much from mpools, as number of requests
allocated is very small. It's less restrictive (mngt_queue does not have
single-CPU affinity) thus avoiding mpool usage in management path allows
to introduce additional restrictions on mpool, leading to I/O performance
improvement.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-09 12:45:51 +02:00
Robert Baldyga
6cd5a27ea9 Merge pull request #812 from Kamoppl/kamilg/fix_tj_actions 
github-actions: update checkout version
 2024-09-09 12:39:31 +02:00
Kamil Gierszewski
0f60163a46 github-actions: update checkout version 
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
 2024-09-09 12:29:01 +02:00
Robert Baldyga
bfd28af1cc Merge pull request #781 from Open-CAS/dependabot/github_actions/dot-github/workflows/tj-actions/changed-files-41 
build(deps): bump tj-actions/changed-files from 18.2 to 41 in /.github/workflows
 2024-09-09 11:19:04 +02:00
Robert Baldyga
193410e7b2 Merge pull request #810 from mmichal10/pt-stats 
Introduce pass-through block stats
 2024-09-06 15:33:23 +02:00
Sara Merzel
835eb708b5 Introduce pass-through block stats 
Signed-off-by: Sara Merzel <sara.merzel@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-06 14:47:02 +02:00
Robert Baldyga
3ebf6e64c1 Merge pull request #808 from mmichal10/vol_and_req_fixes 
Volume improvements
 2024-09-06 14:24:48 +02:00
Robert Baldyga
5a1f6f1f8a Merge pull request #809 from robertbaldyga/ram-needed-no-fail 
Modify ocf_mngt_get_ram_needed to never fail
 2024-09-06 14:20:24 +02:00
Amir Haroush
ed62866324 Modify ocf_mngt_get_ram_needed to never fail 
Signed-off-by: Amir Haroush <amir.haroush@huawei.com>
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
 2024-09-05 15:41:54 +02:00
Gershon Geva
2096e34489 Pass user's params when opening a core volume 
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-04 20:29:41 +02:00
Robert Baldyga
87b16aef6a Do not deinit user volume 
The user is supposed to deinit/destroy it.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-03 16:10:56 +02:00
Robert Baldyga
3d99a2c938 Add missing ocf_volume_init() calls 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-03 16:10:56 +02:00
Robert Baldyga
8aa2d0fb63 Remove unused attach context property 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-03 16:10:56 +02:00
Robert Baldyga
01902e1206 Merge pull request #807 from mmichal10/parallelize-fixes 
Parallelize fixes
 2024-09-03 12:49:54 +02:00
Robert Baldyga
5e6a90a293 parallelize: Fix race condition 
In situation when all the shards finish their work before parallelize
loop does it's final loop condition check, which involves access to
parallelize object, it's possible that parallelize object will be
deinitialized before this final access.

Increasing refcount by 1 before running parallelize and decreasing it
only after the loop is finished addresses this problem.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-02 12:22:11 +02:00
Robert Baldyga
55b99518ed parallelize: Create number of shards requested by user 
In some scenarios running the exact number of shards, regardless of
number of available queues is crucial for correctness of operation.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-02 12:22:11 +02:00
Robert Baldyga
5c714cb3de parallelize: Use mngt_queue only as a fallback 
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
 2024-09-02 12:22:11 +02:00
Robert Baldyga
83d5642547 Merge pull request #806 from mmichal10/deffered-miss 
Count deferred requests as full miss
 2024-09-02 12:18:31 +02:00