Commit Graph

746 Commits

Author SHA1 Message Date
Jan Musial
ae18ce274e Fix cache size requirements and some logging
Signed-off-by: Jan Musial <jan.musial@intel.com>
2022-01-03 14:30:07 +01:00
Robert Baldyga
b40fa0c2bf Fix closing volume on standby stop
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-12-29 20:54:45 +01:00
Robert Baldyga
e73cbad2c7
Merge pull request #631 from mmichal10/dont-stop-cleaner
Don't stop cleaner in activate rollback
2021-12-27 16:51:32 +01:00
Robert Baldyga
0ac66ce4aa Fix cache stop after standby detach
Don't attempt to close cache volume if cache is in standby detached state.

Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-12-23 22:39:37 +01:00
Michal Mielewczyk
a8bdba0cb2 Don't stop cleaner in activate rollback
Activate is not responsible for starting cleaner so rollback shouldn't stop it
eiter.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-12-23 14:46:28 +01:00
Bob Chen
b6de614ada fix volume_close completion order
Signed-off-by: Bob Chen <beef9999@qq.com>
2021-12-22 15:18:34 +08:00
Robert Baldyga
a2916313ee
Revert "fix volume_close completion order" 2021-12-21 20:33:34 +01:00
chenbo
aa6e674034 fix volume_close completion order 2021-12-20 20:10:07 +08:00
Robert Baldyga
0751b2c0c0 Fix metadata flapping
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-12-15 22:21:57 +01:00
Robert Baldyga
cac5869406
Merge pull request #603 from robertbaldyga/metadata-flapping
Introduce flapping of metadata config sections
2021-12-15 17:11:15 +01:00
Robert Baldyga
df9a9f2722 Read superblock sections from cache volume during activate
Because of metadata flapping it is much more complicated to capture those
sections in flight in standby mode, so we read them directly from the cache
volume during the activate.

Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-12-15 15:30:34 +01:00
Robert Baldyga
99c8c05f3f Introduce flapping of metadata config sections
This feature provides double buffering of config sections to prevent
situation when power failure during metadata flush leads to partially
updated metadata. Flapping mechanism makes it always possible to perform
graceful rollback to previous config metadata content in such situation.

Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-12-15 15:30:34 +01:00
Neil Sun
7f82ef3048 Fix incorrect page count calculation with large PAGE_SIZE
e.g., PAGE_SIZE 65536, cache line 8k.

fix https://github.com/Open-CAS/open-cas-linux/issues/1015

Signed-off-by: Sun Feng <loyou85@gmail.com>
2021-12-14 20:07:59 +08:00
Robert Baldyga
60218759d2
Merge pull request #597 from rafalste/fix_core_zero_size_error
Fix core-zero-size error
2021-12-08 22:04:27 +01:00
Robert Baldyga
21c4673251
Merge pull request #600 from mmichal10/cleaning-cmpl
Call completion if failed to perform cleaning
2021-12-08 22:00:58 +01:00
Michal Mielewczyk
d6f2998890 Call completion if failed to perform cleaning
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-12-08 14:50:50 +01:00
Michal Mielewczyk
911a5cddf0 Deinit all registered volume types
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-12-08 14:16:49 +01:00
Michal Mielewczyk
655f732748 Don't access freed memory
Instead of accessing memory of a freed IO, redo size calculations

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-12-08 14:16:49 +01:00
Michal Mielewczyk
244712b020 Prevent race condition in fast path
Request submitted in fast path may be freed before the sequential cutoff stats
are updated. Increment request reference counter to prevent it.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-12-08 09:00:04 +01:00
Rafal Stefanowski
b57bad4652 Fix core-zero-size error
Move error print to where it belongs, preventing this message to
pop up when same error code is reported elsewhere for other reason.

Signed-off-by: Rafal Stefanowski <rafal.stefanowski@intel.com>
2021-12-06 12:30:29 +01:00
Adam Rutkowski
b1494f4642 Remove option to failover without detach
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
2021-11-30 15:18:08 +01:00
Adam Rutkowski
b455a393dd extra assertion in metadata passive update
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
2021-11-30 12:04:57 +01:00
Adam Rutkowski
d0b00817f3 fix cacheline reset in passive metadata update
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
2021-11-30 19:34:52 +01:00
Krzysztof Majzerowicz-Jaszcz
133ea307c8 Fix for issues #988 and #997
This patch fixes the issue 988 (and 997) causing a kernel stack
overflow.

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>
2021-11-24 08:15:07 +01:00
Michal Mielewczyk
4ab22ee2dc Maintain runtime struct during failover standby
To allow the fastest switching from the passive-standby to active mode, the
runtime metadata must be kept 100% synced with the metadata on the drive and in
the RAM thus recovery is required after each collision section update.

To avoid long-lasting recovering of all the cachelines each time the collision
section is being updated, the passive update procedure recovers only those
which have its MD entries on the updated pages.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:58:09 +01:00
Michal Mielewczyk
a6989d1881 Pio concurrency 2021-11-19 11:58:09 +01:00
Michal Mielewczyk
52824adaaf Additional cleaning policy info outside of the SB
Starting cache in a standby mode requires access to a valid cleaning policy
type. If the policy is stored only in the superblock, it may be overridden by
one of the metadata passive updates.

To prevent losing the information it should be stored in cache's runtime
metadata.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
0e529479d6 Init cleaner during passive start
Initializing cleaning policy is very time consuming. To reduce the time required
for activating cache instance the initialization sholud be done during passitve
start

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
390e80794d Refactor cleaning policy initialization
Extract cleaning policy initialization to a separate function

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
6d4e6af5b6 Recovery on passive start
Adjust recovery procedure to allow rebuilding metadata from partialy valid
metadata

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
11dacd6a84 Set dirty shutdown status on standby init
Since part of the recovery is done during `standby init`, the correct shutdown
status has to be set

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
8f58add152 Lru populate unsafe
The unsafe mode is useful if the metadata of added cores is incomplete.

Such scenario is possible when starting cache to standby mode from partially
vaild metadata.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
4deaa1e133 Reset all the status bits during recovery
Make sure all the invalid cachelines have reset status bits. This allows to
recognize invalid cachelines easily during populate.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
fc7c901c8b Skip collision init on cache start passive
Recovery during passive start is based on the assuption that metadata collision
section stored on disk might be partially valid. Reseting this data would make
rebuilding metadata impossible.

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
048bbedd71 Fix metadata_clear_valid_if_clean()
The function should return the cacheline's valid status

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
bb0ff67fe9 Metadata clear_dirty_if_invalid() utility
Fix cacheline's metadata if it is dirty and invalid

Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
800190153b Extend lru list API
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
5ca1404f06 Fix spelling in the error message
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
ccd0abfea5 Add cache line recovery utils to OCF internal API
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Michal Mielewczyk
a7bdaa751d Add error messages on superblock mismatch
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
2021-11-19 11:53:48 +01:00
Rafal Stefanowski
0b39711b8b Add promote-on-threshold sequential cutoff switch
Decide whether to promote sequential cutoff stream
to global structures when threshold is reached

Signed-off-by: Rafal Stefanowski <rafal.stefanowski@intel.com>
2021-11-09 12:54:15 +01:00
Robert Baldyga
55eda53cde
Merge pull request #582 from jfckm/fix-detach-volume
Deinit cache volume on detach
2021-11-09 13:16:08 +01:00
Jan Musial
55eae1447d Deinit cache volume on detach
Signed-off-by: Jan Musial <jan.musial@intel.com>
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
2021-11-08 18:25:30 +01:00
Robert Baldyga
5cbbfdd5ca
Merge pull request #587 from Open-CAS/gcc_11_fix
Fix for OCL issue #968 -  GCC 11 compilation error
2021-11-08 13:34:52 +01:00
Krzysztof Majzerowicz-Jaszcz
99c54be592 Fix for OCL issue #968 - GCC 11 compilation error
GCC 11 static check finds an array size mismatch and prevents OCF from
compiling correctly. This fixes OpenCAS Linux issue #968

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>
2021-11-08 11:39:17 +01:00
Adam Rutkowski
8f080295bb
Merge pull request #585 from rafalste/license_change
Aplying BSD 3-Clause license to OCF source
2021-10-29 13:16:00 +02:00
Rafal Stefanowski
f22da1cde7 Fix license
Change license to BSD-3-Clause

Signed-off-by: Rafal Stefanowski <rafal.stefanowski@intel.com>
2021-10-28 13:08:50 +02:00
Rafal Stefanowski
3cc0d07197 License change to be approved by contributors
Signed-off-by: Rafal Stefanowski <rafal.stefanowski@intel.com>
2021-10-27 12:48:20 +02:00
Adam Rutkowski
12c8b4e333
Merge pull request #574 from Open-CAS/passive_api
Disable selected management operations in failover standby mode
2021-10-25 10:11:26 +02:00
Krzysztof Majzerowicz-Jaszcz
71262d5097 Cache standby mode API changes
Error for an invalid cache operation while in passive mode added

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>

Error name correction

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>

API changes for passive cache mode

Moved the passive cache error return source to the api for flush and
set_param

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>

Further API changes for passive cache mode

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>

 Passive api - review changes

Signed-off-by: Krzysztof Majzerowicz-Jaszcz <krzysztof.majzerowicz-jaszcz@intel.com>
2021-10-22 15:10:53 +02:00