Commit graph

100 commits

Author SHA1 Message Date
d1d123d180 [] writecache: Fix possible panic in initFlushMarks
In case we have many small objects in the write-cache, `indices` should
not be reused between iterations.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2023-02-20 13:53:27 +03:00
e0309e398c [] writecache: Fix possible deadlock
LRU `Peek`/`Contains` take LRU mutex _inside_ of a `View` transaction.
`View` transaction itself takes `mmapLock` [1], which is lifted after tx
finishes (in `tx.Commit()` -> `tx.close()` -> `tx.db.removeTx`)

When we evict items from LRU cache mutex order is different:
first we take LRU mutex and then execute `Batch` which _does_ take
`mmapLock` in case we need to remap. Thus the deadlock.

[1] 8f4a7e1f92/db.go (L708)

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2023-02-20 13:53:27 +03:00
4155c1bdff [] writecache: Use generic LRU cache
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2022-12-31 23:04:06 +03:00
Pavel Karpy
1608fd1c07 [] write-cache: Add "write-cache" to its logs
Signed-off-by: Pavel Karpy <p.karpy@yadro.com>
2022-12-30 11:07:35 +03:00
Pavel Karpy
923f84722a Move to frostfs-node
Signed-off-by: Pavel Karpy <p.karpy@yadro.com>
2022-12-28 15:04:29 +03:00
Evgenii Stratonikov
42554a9298 [] writecache: Remove deleted objects from the writecache
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-12-02 11:52:05 +03:00
Evgenii Stratonikov
4a49ea0855 [] writecache: Allow to open FSTree in read-only mode
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-12-02 11:52:05 +03:00
Evgenii Stratonikov
857d2dc3f5 [] writecache: Optimize initial flush existence checking
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-12-02 11:52:05 +03:00
Pavel Karpy
ed4351aab0 [] write-cache: Do not flush same object twice
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-11-19 11:01:04 +03:00
Pavel Karpy
dd225906a0 [] write-cache: Remove unused variables
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-11-19 11:01:04 +03:00
Evgenii Stratonikov
b0e94b6a6b [] writecache: Do not require read-only mode in Flush
It was needed before we started to flush during transition to
`degraded` mode. Now it is confusing.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-19 11:01:04 +03:00
Evgenii Stratonikov
777fd32d4f [] writecache: Increase error counter on background errors
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-02 14:24:02 +03:00
Evgenii Stratonikov
bffb0f894c [] writecache: Update storage ID during flush
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-02 14:24:02 +03:00
Evgenii Stratonikov
b64b14eb54 [] writecache: Reuse FSTree flushing code between flushes
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-02 14:24:02 +03:00
Evgenii Stratonikov
a56927e3d4 [] writecache: Remove unused variable
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-02 14:24:02 +03:00
Evgenii Stratonikov
98a152256b [] writecache: Allow to open in NOSYNC mode
Applicable only to FSTree as we cannot handle corrupted databases
properly yet.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-11-01 09:42:26 +03:00
Evgenii Stratonikov
34501685b7 [] local_object_storage: Move ErrObjectIsExpired to another package
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-10-31 11:41:24 +03:00
Evgenii Stratonikov
56de2f1363 [] local_object_storage: Simplify logic error construction
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-10-31 11:41:24 +03:00
Evgenii Stratonikov
fcdbf5e509 [] local_object_storage: Add a type for logical errors
All logic errors are wrapped in `logicerr.Logical` type and do not
affect shard error counter.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-10-31 11:41:24 +03:00
Leonard Lyubich
e8c5f03c30 [] shard: Don't log read-only errors of write-cache
There is no need to log `writecache.ErrReadOnly` errors in `Delete`
method of the `Shard`.

Signed-off-by: Leonard Lyubich <ctulhurider@gmail.com>
2022-10-28 18:30:45 +03:00
Pavel Karpy
f037022a7a [] logger: Refactor Logger component
Make it store its internal `zap.Logger`'s level. Also, make all the
components to accept internal `logger.Logger` instead of `zap.Logger`; it
will simplify future refactor.

Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-10-12 18:11:05 +03:00
Evgenii Stratonikov
8b3b16fe62 [] writecache: Flush cache when moving to the DEGRADED mode
Degraded mode allows us to operate without an SSD,
thus writecache should be unavailable in this mode.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-10-04 12:13:09 +03:00
Evgenii Stratonikov
0a411908ee [] writecache: Allow to ignore read errors during flush
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-28 09:28:01 +03:00
Evgenii Stratonikov
0b4c867ef1 [] writecache: Allow to start flush manually
Allow user to initiate flushing objects from a writecache.
We need this in 2 cases:
1. During writecache storage schema update, it should be flushed with
   the old version of node and started clean with a new one.
2. During SSD replacement, to avoid data loss.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-28 09:28:01 +03:00
Evgenii Stratonikov
55148404ee [] writecache: Use interfaces for blobstor and metabase
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-28 09:28:01 +03:00
Evgenii Stratonikov
e6cf0b3d42 [] writecache: Set flush parameters based on max size
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
c04126f35e [] writecache: Remove IsErrNotFound
We specify the error in the doc-comment, and it is the same for all our
components.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
9e41e85295 [] writecache: Actualize docs
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
e9c6ee2623 [] writecache: Rename constants for default values
Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
177e8e01b1 [] neofs-node: Remove memcache_capacity from the configuration
It is unused since ddaed283e9 .

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
bda084f331 [] writecache: Simplify object counters
Remove unused option and additional pointers to db/fstree.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
20abdaeed4 [] writecache: Set flush mark only on success
Set flush mark in the inside the flush worker because writing to the blobstor
can fail. Because each evicted object must be deleted, it is reasonable
to do this in the evict callback.

The evict callback is protected by LRU mutex and thus potentially interferes
with `Get` and `Iterate` methods. This problem will be addressed in the
future.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-09-02 11:59:53 +03:00
Evgenii Stratonikov
b8b9d25f9d [] local_object_storage: Add generic tests
Use them for writecache as a simple example.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-08-30 10:02:43 +03:00
Pavel Karpy
30341f2192 [] *: Perform go fmt using go v1.19
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-08-22 18:59:57 +03:00
Evgenii Stratonikov
26b4a258e0 [] neofs-node: Refactor configuration
1. Move compression parameters to the `shard` section.
2. Allow to use multiple sub-storage components in the blobstor.

Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
266458fe5c [] blobstor: Unify fstree and blobovnicza interfaces
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
5aa3defc67 [] writecache: Remove cloneBytes
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
ddaed283e9 [] writecache: Simplify logic
1. Remove in-memory cache. It doesn't persist objects and if we want
   more speed, `NoSync` option can be used for the bolt DB.
2. Put to the metabase in a synchronous fashion. This considerably
   simplifies overall logic and plays nicely with the metabase bolt DB
   batch settings.

Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
4176b2a1bc [] local_object_storage: Unify parameters for the Iterate operation
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
73f8bb3e5f [] shard: Store generic storage ID in metabase
Allow to extend blobstor with more storage sub-systems. Currently
objects stored in the FSTree have empty byte slice descriptor and object
from blobovnicza tree have the same id as earlier. Each such change in
the identifier formation should be accompanied with metabase version
increase.

Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
9eb018672c [] blobstor: Unify parameters for Put operation
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
ca15083a50 [] local_object_storage: Unify parameters for the Exists operation
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
d9a2f280c9 [] local_object_storage: Unify parameters for the Get operation
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-08-22 13:14:19 +03:00
Evgenii Stratonikov
4e361e2ab4 [] writecache: Open DB in sync mode
Close .

Initially, `Sync: false` was provided because we can already lose
objects cached in memory. However, future changes in writecache will
remove inmemory cache and speed up it via other means.

Signed-off-by: Evgenii Stratonikov <evgeniy@morphbits.ru>
2022-08-16 15:27:18 +04:00
Pavel Karpy
da2975a2f9 [] write-cache: Fix panic on Delete operation
If an object is found in the Write-cache and is placed at the end of
the in-memory cache, the memory counter update operation tries to
dereference the index that is out of the sliced array. Moreover, even if
panic does not appear, the counter is updated with the wrong value.

Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-08-09 20:18:26 +03:00
Pavel Karpy
156ba85326 [] node: Do not return expired objects
If an object has not been marked for removal by the GC in the current epoch
yet but has already expired, respond with `ErrObjectNotFound` api status.
Also, optimize shard iteration: a node must stop any iteration if the object
 is found but gonna be removed soon.
All the checks are performed by the Metabase.

Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
2022-08-04 16:31:49 +03:00
Evgenii Stratonikov
1691364653 [] local_object_storage: Fix tests and some data races
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-07-21 17:56:06 +03:00
Evgenii Stratonikov
3df62769c0 [] local_object_storage: Allow to set mode for all components
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-07-21 17:56:06 +03:00
Evgenii Stratonikov
1e786233bf [] local_object_storage: Provide readOnly flag to Open
We should be able to reopen storage in readonly in runtime.

Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-07-21 17:56:06 +03:00
Evgenii Stratonikov
339864b720 [] local_object_storage: Move shard.Mode to a separate package
Signed-off-by: Evgenii Stratonikov <evgeniy@nspcc.ru>
2022-07-21 17:56:06 +03:00