Aliaksandr Valialkin
6865f3b497
Revert "lib/mergeset: remove a pool for inmemoryBlock structs"
...
This reverts commit 793fe39921
.
Reason to revert: production testing revealed possible slowdown when registering big number of new time series
2021-05-28 01:11:22 +03:00
Aliaksandr Valialkin
7b33bc67a1
lib/mergeset: remove a pool for inmemoryBlock structs
...
The pool for inmemoryBlock struct doesn't give any performance gains in production workloads,
while it may result in excess memory usage for inmemoryBlock structs inside the pool during
background merge of indexdb.
2021-05-27 22:00:50 +03:00
Aliaksandr Valialkin
0fc857d363
lib/{mergeset,storage}: reduce the number of IFNO log messages like merged ... items across ... blocks in ... seconds
...
Log these messages if the merge takes more than 30 seconds instead of 10 seconds.
2021-05-23 14:15:49 +03:00
Aliaksandr Valialkin
e37e1b1e34
lib/{storage,mergeset}: fix unaligned 64-bit atomic operation
panic for 32-bit architectures
...
The panic has been introduced in 56b6b893ce
2021-04-27 16:42:19 +03:00
Aliaksandr Valialkin
2d1d60118d
lib/mergeset: split rows ingestion among multiple shards
...
This improves rows ingestion on systems with many CPU cores by reducing lock contention.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1244
Thanks to @waldoweng for the original idea and draft implementation at https://github.com/VictoriaMetrics/VictoriaMetrics/pull/1243
2021-04-27 15:45:11 +03:00
Aliaksandr Valialkin
ab8008d6d7
lib/{storage,mergeset}: remove empty directories on startup. Such directories can be left after unclean shutdown on NFS storage
...
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/1142
2021-04-22 13:03:29 +03:00
Aliaksandr Valialkin
587132555f
lib/mergeset: reduce memory usage for inmemoryBlock by using more compact items representation
...
This also should reduce CPU time spent by GC, since inmemoryBlock.items don't have pointers now,
so GC doesn't need visiting them.
2021-02-21 22:09:10 +02:00
Aliaksandr Valialkin
b8a5ee2e93
lib/{mergeset,storage}: allow merging smaller number of small parts
...
While this may increase CPU and disk IO usage needed for background merge,
this also recudes CPU usage during queries in production. This is because
such queries tend to read recently added data and it is better to have lower number
of parts for such data in order to reduce CPU usage.
This partially reverts ebf8da3730
2021-02-21 21:43:37 +02:00
Aliaksandr Valialkin
34195218e1
lib/{mergeset,storage}: do not use pools for indexBlock and inmemoryBlock during their caching, since this results in higher memory usage in production without any performance gains
2021-02-21 21:43:37 +02:00
Aliaksandr Valialkin
9566015a36
Revert "lib/mergeset: tune lifetime for entries inside block caches"
...
This reverts commit 458c89324d
.
Production testing revealed zero improvements for memory usage with reduced lifetime for entries in block caches.
2021-02-17 20:42:15 +02:00
Aliaksandr Valialkin
35a23234ca
lib/mergeset: tune lifetime for entries inside block caches
...
This should reduce memory usage in general case without significant CPU usage increase
2021-02-16 18:12:32 +02:00
Aliaksandr Valialkin
500acb958c
lib/mergeset: clarify comments in the code a bit
2021-02-16 18:03:11 +02:00
Aliaksandr Valialkin
d788876a7a
lib/mergeset: remove unused code after a4140de9e6
2021-02-16 13:40:29 +02:00
Aliaksandr Valialkin
afa9cf9c57
lib/mergeset: remove dead code left after a4140de9e6
2021-02-09 16:51:09 +02:00
Aliaksandr Valialkin
7b7963a77f
lib/mergeset: unconditionally cache indexdb blocks
...
Production workloads show that indexdb blocks must be cached unconditionally for reducing CPU usage.
This shouldn't increase memory usage too much, since unused blocks are removed from the cache every two minutes.
2021-02-09 00:49:59 +02:00
Aliaksandr Valialkin
e8ee9fa7fe
app/vmstorage: export missing vm_cache_size_bytes
metrics for indexdb and data caches
2021-02-09 00:49:58 +02:00
faceair
cf76d8dd79
lib/mergeset: add missing shouldCacheBlock ( #1019 )
2021-01-15 11:47:25 +02:00
Aliaksandr Valialkin
edbe35509e
lib/{storage,mergeset}: tune background merge process in order to reduce CPU usage and disk IO usage
2020-12-18 20:01:20 +02:00
Aliaksandr Valialkin
1a237c6903
all: properly handle CPU limits set on the host system/container
...
This can reduce memory usage on systems with enabled CPU limits.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/946
2020-12-08 21:07:03 +02:00
Aliaksandr Valialkin
8f8727cb65
lib/mergeset: tune the number of rawItemsBlocks to merge at once
...
512 blocks give higher ingestion performance and slightly lower memory usage
2020-11-25 21:53:15 +02:00
Aliaksandr Valialkin
8fcd87a6a5
lib/mergeset: help GC by removing refereces to slices in inmemoryBlock.Reset
2020-11-25 21:20:02 +02:00
Aliaksandr Valialkin
7f3e884a31
all: spelling fix: superflouos->superfluous. This is a follow-up for 0acdab3ab9
2020-11-24 12:42:04 +02:00
Aliaksandr Valialkin
f4fd917e4f
lib/fs: replace fs.OpenReaderAt with fs.MustOpenReaderAt
...
All the callers for fs.OpenReaderAt expect that the file will be opened.
So it is better to log fatal error inside fs.MustOpenReaderAt instead of leaving this to the caller.
2020-11-23 09:57:30 +02:00
Aliaksandr Valialkin
c736339843
lib/{storage,mergeset}: clean cached index blocks and inmemory blocks more aggressively
...
Previously such blocks were cleaned after they weren't accessed during 10 minutes.
Now they are cleaned after one minute of missing access. This should reduce memory usage in general case.
2020-11-04 16:44:15 +02:00
Aliaksandr Valialkin
3abbb38254
lib/{mergeset,storage}: compare errors with errors.Is()
2020-09-17 03:03:10 +03:00
Aliaksandr Valialkin
ddb3519e17
lib/{mergeset,storage}: code prettifying
2020-09-17 02:06:37 +03:00
Aliaksandr Valialkin
edf3ee2a9b
lib: dump compressed block contents on error during decompression
...
This should improve detecting root cause for https://github.com/facebook/zstd/issues/2222
2020-08-15 14:51:14 +03:00
Aliaksandr Valialkin
29bbab0ec9
lib/storage: remove prioritizing of merging small parts over merging big parts, since it doesn't work as expected
...
The prioritizing could lead to big merge starvation, which could end up in too big number of parts that must be merged into big parts.
Multiple big merges may be initiated after the migration from v1.39.0 or v1.39.1. It is OK - these merges should be finished soon,
which should return CPU and disk IO usage to normal levels.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/648
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/618
2020-07-30 20:02:22 +03:00
Aliaksandr Valialkin
b8303afcd8
lib/storage: improve prioritizing of data ingestion over querying
...
Prioritize also small merges over big merges.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/291
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/648
2020-07-23 01:40:38 +03:00
Aliaksandr Valialkin
6afdcf8a20
lib/mergeset: properly calculate global metrics in UpdateStats()
...
Previously these metrics could be calculated multiple times for multiple mergeset.Table instances.
2020-07-23 00:35:29 +03:00
Aliaksandr Valialkin
d962568e93
all: use %w instead of %s for wrapping errors in fmt.Errorf
...
This will simplify examining the returned errors such as httpserver.ErrorWithStatusCode .
See https://blog.golang.org/go1.13-errors for details.
2020-06-30 23:33:46 +03:00
Aliaksandr Valialkin
a72f18e821
lib/{storage,mergeset}: further tuning of compression levels depending on block size
...
This should improve performance for querying newly added data, since it can be unpacked faster.
2020-05-15 13:12:28 +03:00
Aliaksandr Valialkin
6838fa876c
lib/mergeset: tune compression levels in order to improve ingestion performance a bit
2020-05-15 12:12:15 +03:00
Aliaksandr Valialkin
3845420a8f
lib: extract common code for returning fast unix timestamp into lib/fasttime
2020-05-14 23:06:50 +03:00
Aliaksandr Valialkin
7e831741f9
lib/{storage,mergeset}: return dst on error from unmarshalBlockHeaders, so it could be reused
2020-05-14 15:32:23 +03:00
Aliaksandr Valialkin
f442d81648
lib/{storage,mergeset}: cleanup: remove unused partSearch.indexBlockReuse
2020-05-14 14:03:15 +03:00
Aliaksandr Valialkin
0b2f678d8e
lib/{storage,mergeset}: make sure that requests
and misses
cache counters never go down
2020-04-10 14:44:52 +03:00
Aliaksandr Valialkin
e3b18ca1ab
lib/mergeset: skip createing temporary part objects when merging source inmemory parts
...
This should reduce CPU usage when adding new entries to inverted index.
This should alos prevent from creating stalled cleaner goroutines for the created temporary parts,
since they were never closed.
This should fix the following issue: https://github.com/VictoriaMetrics/VictoriaMetrics/issues/316 .
2020-02-13 14:09:13 +02:00
Aliaksandr Valialkin
347aaba79d
lib/{storage,mergeset}: use time.Ticker instead of time.Timer where appropriate
...
It has been appeared that time.Timer was used in places where time.Ticker must be used instead.
This could result in blocked goroutines as in the https://github.com/VictoriaMetrics/VictoriaMetrics/issues/316 .
2020-02-13 13:21:48 +02:00
Aliaksandr Valialkin
da19fffa08
all: rename ReadAt* to MustReadAt* in order to dont clash with io.ReaderAt
2020-01-30 15:16:16 +02:00
Aliaksandr Valialkin
cb2a2f281f
lib/mergeset: properly update lastAccesstime
in indexBlockCache entries
...
This is a follow-up for 6665f10e7b
2020-01-29 21:21:01 +02:00
Aliaksandr Valialkin
ea53a21b02
all: consistently log durations in seconds with millisecond precision
...
This should improve logs readability
2020-01-22 18:35:24 +02:00
Aliaksandr Valialkin
62b041e90a
lib/{mergeset,storage}: properly update lastAccessTime
in index and data block cache entries
2020-01-20 15:00:10 +02:00
Aliaksandr Valialkin
caffb0cd01
lib/{mergeset,storage}: fix uint64 counters alignment for 32-bit architectures (GOARCH=386, GOARCH=arm)
2020-01-14 22:47:42 +02:00
Aliaksandr Valialkin
b03ccbf6f7
lib/{storage,mergeset}: gradually remove stale entries from block cache and index caches
...
This should reduce memory usage in the long run when old blocks and indexes
aren't accessed anymore.
2020-01-14 21:38:29 +02:00
Aliaksandr Valialkin
5d2ff573aa
app/vmselect/promql: allow negative offsets
...
Updates https://github.com/prometheus/prometheus/issues/6282
2019-12-11 00:57:51 +02:00
Aliaksandr Valialkin
3694efd005
lib/{mergeset,storage}: log info message when both source and destination part paths from txn are missing during startup
...
This is expected condition after unclean shutdown (OOM, hard reset, `kill -9`) on NFS disk.
2019-12-09 15:45:23 +02:00
Aliaksandr Valialkin
639967db59
lib/{mergeset,storage}: make sure pending transaction deletions are finished before and after runTransactions
call.
...
`runTransactions` call issues async deletions for transaction files. The previously issued transaction deletions
can race with the next call to `runTransactions`. Prevent this by waiting until all the pending transaction
deletions are funished in the beginning of `runTransactions`. Also make sure that all the pending transaction
deletions are finished before returning from `runTransactions`.
2019-12-04 21:40:52 +02:00
Aliaksandr Valialkin
b9616c017f
lib/{mergeset,storage}: remove transaction files only after the mentioned dirs are really removed
...
This should fix the issue on NFS when incompletely removed dirs may be left
after unclean shutdown (OOM, kill -9, hard reset, etc.), while the corresponding transaction
files are already removed.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/162
2019-12-02 21:34:37 +02:00
Aliaksandr Valialkin
26ffc77622
lib/{storage,mergeset}: create missing partition directories after restoring from backups
...
Backup tools could skip empty directories. So re-create such directories on the first run.
2019-11-02 02:27:19 +02:00