Commit graph

1203 commits

Author SHA1 Message Date
b142b6f48e [#1367] fstree: Add size to file counter
FSTree file counter used by writecache. As writecache has now only one
storage, so it is required to use real object size to get writecache
size more accurate than `count * max_object_size`.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-12 15:06:33 +03:00
5f6c7cbdb1 [#1367] writecache: Drop bbolt DB
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-12 14:22:29 +03:00
92fe5d90f5 [#1359] writecache: Drop unused
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-09 18:45:21 +03:00
4668efc0bf [#1355] metabase: Upgrade improvements
Do not fail on same latest version to run compact on upgraded metabase.
Use NoSync on compact.
Log every batch on bucket delete stage.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-09 11:37:36 +03:00
d3b209c8e1 [#1337] shard: Disable background rebuild
Since `frostfs-cli control shards rebuild` command was added,
there is no need for background rebuild now.
For failover tests used used value 1 to rebuild only schema change.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 15:19:55 +03:00
edb1747af7 [#1337] blobovniczatree: Add rebuild by overflow
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 15:19:54 +03:00
a61201a987 [#1337] config: Move rebuild_worker_count to shard section
This makes it simple to limit performance degradation for every shard
because of rebuild.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 13:57:27 +03:00
6b6eabe41c [#1337] cli: Add control shards rebuild command
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 13:57:27 +03:00
d508da8397 [#1337] blobovniczatree: Add rebuild by fill percent
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 13:57:27 +03:00
007827255e [#1337] blobovniczatree: Add .rebuild temp files
This allows to reduce open/close DBs to check incompleted rebuilds.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-09-06 13:57:27 +03:00
108e4e07be [#1349] node: Evacuate objects without setting mode to MAINTENANCE
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-09-05 16:08:27 +03:00
a685fcdc96 [#1317] go.mod: Use range over int
Since Go 1.22 a "for" statement with a "range" clause is able
to iterate through integer values from zero to an upper limit.

gopatch script:
@@
var i, e expression
@@
-for i := 0; i <= e - 1; i++ {
+for i := range e {
    ...
}

@@
var i, e expression
@@
-for i := 0; i <= e; i++ {
+for i := range e + 1 {
    ...
}

@@
var i, e expression
@@
-for i := 0; i < e; i++ {
+for i := range e {
    ...
}

Signed-off-by: Ekaterina Lebedeva <ekaterina.lebedeva@yadro.com>
2024-09-03 13:00:54 +03:00
2b3fc50681
[#1320] shard: Fix TestGCDropsObjectInhumedFromWritecache flaky test
The `TestGCDropsObjectInhumedFromWritecache` test was flaky because a
running asynchronous rebuild operation prevented GC from deleting the
object. A test-only shard option `WithDisabledRebuild` has been added
to fix this.

Signed-off-by: Aleksey Savchuk <a.savchuk@yadro.com>
2024-09-02 10:26:53 +03:00
882c068410 [#1334] metabase: Store upgrade flag
This allows to check if metabase upgrade was not completed.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-30 09:06:20 +03:00
6c2146bbc1 [#1334] metabase: Add upgrade from v2 to v3
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-30 09:06:20 +03:00
03976c6ed5 [#1341] .golangci.yml: Replace exportloopref with copyloopvar
exportloopref is deprecated.
gopatch:
```
@@
var index, value identifier
var slice expression
@@
for index, value := range slice {
...
-value := value
...
}

@@
var index, value identifier
var slice expression
@@
for index, value := range slice {
...
-index := index
...
}

@@
var value identifier
var channel expression
@@
for value := range channel {
...
-value := value
...
}
```

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-08-28 15:44:41 +00:00
806ea37101 [#1328] pilorama: Do not skip items in SortedByFilename
Benchmark results:
```
goos: linux
goarch: amd64
pkg: git.frostfs.info/TrueCloudLab/frostfs-node/pkg/local_object_storage/pilorama
cpu: 11th Gen Intel(R) Core(TM) i5-1135G7 @ 2.40GHz
                                   │     old     │                 new                 │
                                   │   sec/op    │   sec/op     vs base                │
ForestSortedIteration/bbolt,root-8   207.2µ ± 6%   173.6µ ± 6%  -16.23% (p=0.000 n=10)
ForestSortedIteration/bbolt,leaf-8   3.910µ ± 5%   3.928µ ± 7%        ~ (p=0.529 n=10)
geomean                              28.46µ        26.11µ        -8.27%
```

They are not representative, as the worst case is when we have multiple
items of different lengths. However, `FileName` is usually less than 100
in practice, so the asymptotics is the same.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-08-26 06:11:32 +00:00
80099d9a2f [#1328] pilorama: Add tricky test for SortedByFilename
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-08-26 06:11:32 +00:00
76f67ea34e [#1323] metabase: Bump version
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
7d0d781db1 [#1323] metabase: Drop user attribute index
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
0f08a2efba [#1323] metabase: Resolve funlen linter
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
7bf20c9f1f [#1323] metabase: Add expiration epoch buckets
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
2542d4f5df [#1323] metabase: Drop payload checksum index
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
15dae8685e [#1323] metabase: Drop ownerID index
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-22 08:21:40 +00:00
7bca428db0 [#1322] Use new protobuf marshaler
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-08-22 07:17:41 +00:00
a345c972bf [#1316] lint: Fix warnings
Renamed parameters `min/max` to avoid conflicts with
predeclared identifiers.

Replaced background context with parent context without
cancellation in closer functions in frostfs-node.

Signed-off-by: Ekaterina Lebedeva <ekaterina.lebedeva@yadro.com>
2024-08-21 18:18:25 +03:00
cfda9003a7 [#1318] meta: Add test TestInhumeECObject
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-08-20 16:21:27 +03:00
6ff0b0996b [#1318] metrics: Fix container_size_bytes for EC
When node put chunk into EC container, `policer` may remove it as redundant.
This chunk marked as removed. When parent object removed and `gc` start iterating over chunk,
node count removing chunk twice.

Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-08-20 15:33:28 +03:00
80ce7c3a00 [#1284] shard: Resolve funlen linter
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-09 13:23:35 +03:00
93d63e1632 [#1284] writecache: Allow to seal writecache async
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-09 13:23:35 +03:00
68029d756e [#1302] writecache: Allow to specify custom page size
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-09 12:12:07 +03:00
fa82854af4 [#1302] writecache: Add put->flush->put benchmark
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-09 12:12:07 +03:00
08b1f18bca [#1296] writecache: Add count limit
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-09 06:30:32 +00:00
36efccd862 [#1298] writecache: Add shrink flag for Seal command
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-08 16:32:29 +03:00
5c01bd5be8 [#1298] writecache: Add restore-mode flag for Seal command
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-07 11:20:14 +03:00
8e51d7849a [#1295] getSvc: Assemble complex EC object headers without linking object
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-06 16:48:12 +03:00
10602b55b1 [#1295] engine: Resolve funlen linter
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-06 09:45:47 +03:00
eeca796d2e [#1295] engine: Log object address in case of error
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-08-06 09:45:43 +03:00
18182e578e [#1261] shard: Fix delete objects from FSTree
Replace nil storageID with empty like by shard.Get.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-22 14:38:09 +03:00
3119f2fbc3 [#1257] metabase: Delete EC gc marks and split info
EC parent and split gc marks should be deleted after last EC chunk delete.
Also should delete split info for EC parent and split parent.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-19 17:14:14 +03:00
27caa8a72f [#1256] metabase: Put split parent object ID for EC chunks
It is required to save split parent ID too, not only split ID.
Otherwise inhume operation works incorrect: shard with last part may be skipped
and parent object will be available.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-17 18:08:07 +03:00
3940bc17c1 [#1251] pilorama: Allow traversing multiple branches in parallel
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-07-17 11:25:07 +03:00
3bf6e6dde6 [#1238] engine: Add non-blocking send in the shard's notification channel
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-07-10 11:37:11 +03:00
b027a7f91e [#1234] pilorama: Fix GetByPath() on duplicate directories
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-07-10 06:11:38 +00:00
6ace2f597e [#1234] pilorama: Add test for duplicate directory behaviour
When AddByPath() is called concurrently on 2 different nodes,
internal path components may be created twice. This violates some
of our assumptions in GetByPath() and, indirectly, in S3 handling of
GetSubTree() results.

Add a test for the correct behaviour, fixes will follow.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-07-10 06:11:38 +00:00
62cbb72a5e [#1226] blobovniczatree: Delete fix db extensions in Init()
Since several releases have been released, this code is not relevant.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-04 12:22:06 +03:00
78b1d9b18d [#1226] blobovniczatree: Drop leaf width limitation
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-04 12:22:06 +03:00
40c9ddb6ba [#1226] blobovniczatree: Drop init in advance option
To make blobovniczatree unlimited.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-04 12:22:06 +03:00
3a797e4682 [#1222] engine: Fix tree evacuation
Do not fail if it is unable to evacuate tree to other node.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-04 10:38:10 +03:00
2bac82cd6f [#1222] engine: Fix object evacuation
Do not fail evacuation if it unable to evacuate object to other node.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-04 10:38:10 +03:00
bbe95dac8b [#1225] engine: Log the error when check object existence
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-07-04 07:22:54 +00:00
56eeb630b6 [#1217] Fix grammar mistakes and misspelling
Signed-off-by: Ekaterina Lebedeva <ekaterina.lebedeva@yadro.com>
2024-07-01 19:14:25 +03:00
7a8ac4907a [#1213] engine: Drop unused
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-07-01 06:49:35 +00:00
7085723c6b [#1074] pilorama: Allow empty filenames in SortedByFilename()
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-06-28 17:46:24 +03:00
4c7ff159ec [#1201] writecache: Cancel background flush without lock
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-28 09:02:36 +03:00
4951babd5f [#1208] blobstor: Fix delete without storage id
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-27 11:59:08 +03:00
81ea91de52 [#451] metrics: Move to internal
`metrics` don't look like something others want to import.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-25 08:52:37 +00:00
9ac74efc41 [#1173] shard: Use mode from config on reload
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-06-20 11:29:10 +00:00
75eedf71f3 [#1187] pilorama/test: Remove debug print
Introduced in e12fcc041d.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-06-18 15:09:44 +03:00
76cf7a051b [#1178] shard: Check metabase existence before read shard ID
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-17 09:59:15 +03:00
68ac490729 [#1174] shard: Update metric mode_info on Init
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-06-13 08:32:59 +00:00
6a39c3d15e [#1086] engine: Do not use metabase if shard looks bad
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-13 07:35:22 +00:00
9d73f9c2c6 Reapply "[#446] engine: Move to read-only on blobstor errors"
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-06-13 07:35:22 +00:00
b9fcaad21f [#1168] shard: Set Disabled as default mode for components
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-06-11 15:13:38 +00:00
6cf512e574 [#1166] blobovniczatree: Handle blobovnicza's NoSpaceLeft error
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-07 17:15:43 +03:00
e7d479f4c2 [#1166] blobovnicza: Return NoSpaceLeft error instead of syscall.ENOSPC
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-07 17:15:43 +03:00
3f1961157e [#1163] metabase: Handle multiple splitInfos for EC
For REP updating split info is handled explicitly by a high-level PUT logic.
For EC it is trickier, because the address of an object we put is only
distantly related to a split info.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-06-06 16:26:29 +03:00
2e074d3846 [#1163] metabase: Properly save EC parent split ID
Search by SplitID should return all parts of a complex object.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-06-05 12:40:16 +03:00
806236da78 [#1121] node: Change mode of shard components
Signed-off-by: Alexander Chuprov <a.chuprov@yadro.com>
2024-06-05 05:55:24 +00:00
6f2187a420 [#1121] node: Refactor mods of shard
Signed-off-by: Alexander Chuprov <a.chuprov@yadro.com>
2024-06-05 05:55:24 +00:00
cc2449beaf [#1158] metabase: Fix EC storage schema
Do not store EC info twice.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-06-04 17:24:40 +03:00
5aacb8fc86 [#1144] metabase: Save parent attributes for ec-chunks
Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
2024-05-31 19:55:32 +03:00
f8e33f8e3a [#1144] metabase: Proprely choose root OID for EC-splitted objects
* If EC-parent is a part of Split itself, then save to root bucket
  its parent;
* If EC-parent is not a part of Split itself, then save to root bucket
  OID of this EC-parent.

Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
2024-05-31 19:53:32 +03:00
f0edebea18 [#1144] metabase: Support ec parent filter for Search
Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
2024-05-31 19:53:32 +03:00
92e19feb57 [#1147] node: Use public fields for shard.ExistsPrm
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-05-30 08:13:04 +00:00
6130650bb6 [#1147] node: Implement Lock\Delete requests for EC object
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-05-30 08:13:04 +00:00
a82c8cc5b8 [#1147] gc: Execute callback for expired tombstones when they exists
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-05-30 08:13:04 +00:00
40b04c00ef [#1141] metabase: Fix IsUserObject method
Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
2024-05-20 13:24:15 +03:00
89a80e9a0f [#1141] metabase: Fix putUniqueIndexItem
* `GetECHeader` is not correct way to determine if an object's got
  EC-header: `ECHeader` must be used for that.

Signed-off-by: Airat Arifullin <a.arifullin@yadro.com>
2024-05-20 13:24:08 +03:00
d45d086acd [#1129] policer: Add EC chunk replication
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-05-16 16:28:48 +03:00
5c582e96fd [#1136] metabase: Fix creation of ECInfoError
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-05-15 11:04:27 +00:00
6e71ae3bda [#1130] fstree: Remove useless Stat() call
```
goos: linux
goarch: amd64
pkg: git.frostfs.info/TrueCloudLab/frostfs-node/pkg/local_object_storage/blobstor
cpu: 11th Gen Intel(R) Core(TM) i5-1135G7 @ 2.40GHz
                                           │     old     │                new                 │
                                           │   sec/op    │   sec/op     vs base               │
SubstorageReadPerf/fstree_nosync-seq100-8    2.689µ ± 2%   2.428µ ± 4%  -9.72% (p=0.000 n=10)
SubstorageReadPerf/fstree_nosync-rand100-8   2.727µ ± 1%   2.497µ ± 2%  -8.42% (p=0.000 n=10)
geomean                                      2.708µ        2.462µ       -9.07%
```

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-05-14 16:05:45 +03:00
00b2b77b26 [#1112] node: Implement Range\RangeHash requests for EC object
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-05-07 14:47:21 +03:00
e07869a8cf [#1100] Remove unused fields
Signed-off-by: Ekaterina Lebedeva <ekaterina.lebedeva@yadro.com>
2024-05-06 10:14:36 +03:00
c9efaa5819 [#966] node: Add path of the write_cache to metric labels
Signed-off-by: Alexander Chuprov <a.chuprov@yadro.com>
2024-05-02 06:46:46 +00:00
4730ecfdb8 [#966] node: Refactor WriteCacheMetrics interface
Grouping common fields of methods will enhance the readability of the interface.

Signed-off-by: Alexander Chuprov <a.chuprov@yadro.com>
2024-05-02 06:46:46 +00:00
411a8d0245 [#1004] blobovnicza: Use TTL for blobovnicza tree cache
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-04-26 19:54:29 +03:00
112a7c690f [#1103] node: Implement Get\Head requests for EC object
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-04-24 18:15:53 +03:00
167c52a1a9 [#1103] node: Reduce amount of lines for method StorageEngine.head
Signed-off-by: Anton Nikiforov <an.nikiforov@yadro.com>
2024-04-24 16:31:04 +03:00
e5e0542482 [#1085] log: Move storage log message to constants package
Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-15 07:57:00 +00:00
5be36924e3 [#41] log: Log storage operations in only in Debug
They are mostly useless unless we need to _debug_ a specific issue.
The amount of logs we produce is too big.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-15 07:57:00 +00:00
40781b3a20 [#1086] engine: Change mode in case of errors async
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-04-10 12:29:43 +00:00
5ef5734c4e Reapply "[#972] Drop x/exp/slices dependency"
This reverts commit 946f2ec2bf.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-10 12:09:34 +00:00
2c4b50a71e Reapply "[#972] Use require.ElementsMatch() where possible"
This reverts commit 7627d08914.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-10 12:09:34 +00:00
3dc81cb4fc Reapply "[#972] Use min/max builtins"
This reverts commit dad56d2e98.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-10 12:09:34 +00:00
d864945961 Reapply "[#972] pilorama: Remove removeDuplicatesInPlace()"
This reverts commit 9f68305c2e.

Signed-off-by: Evgenii Stratonikov <e.stratonikov@yadro.com>
2024-04-10 12:09:34 +00:00
1b17258c04 [#1029] metabase: Add refill metrics
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-04-10 13:05:44 +03:00
e3d9dd6ee8 [#1024] blobovnicza: Copy data on iterate
DB value is only valid while the tx is alive.
But handler may to run something in other goroutine.

Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-04-10 10:21:11 +03:00
57466594fb [#1024] shard: Resync metabase concurrently
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-04-10 10:21:10 +03:00
1005bf4f56 [#1024] shard: Add refill metabase benchmark
Signed-off-by: Dmitrii Stepanov <d.stepanov@yadro.com>
2024-04-10 10:21:10 +03:00