* feat: add option to flush WAL on shutdown
Add `--storage-wal-flush-on-shutdown` to flush WAL on database shutdown.
On successful shutdown, all WAL data will be committed to TSM files and the
WAL directories will not contain any .wal files.
Closes: #25422
* feat(tsm1/wal): encapsulate expiring WAL files in FileDisposer
This changeset introduces an interface extension point named
FileDisposer to control what to do with WAL files when they are no
longer needed. Currently, the only implementation is to delete the file
which is the existing behavior.
* chore: accumulate errors
Since we're here, capture the previously ignored fs errors and pass up a
combined error (which the only callers log out).
* test: use `T.TempDir` to create temporary test directory
This commit replaces `os.MkdirTemp` with `t.TempDir` in tests. The
directory created by `t.TempDir` is automatically removed when the test
and all its subtests complete.
Prior to this commit, temporary directory created using `os.MkdirTemp`
needs to be removed manually by calling `os.RemoveAll`, which is omitted
in some tests. The error handling boilerplate e.g.
defer func() {
if err := os.RemoveAll(dir); err != nil {
t.Fatal(err)
}
}
is also tedious, but `t.TempDir` handles this for us nicely.
Reference: https://pkg.go.dev/testing#T.TempDir
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: fix failing TestSendWrite on Windows
=== FAIL: replications/internal TestSendWrite (0.29s)
logger.go:130: 2022-06-23T13:00:54.290Z DEBUG Created new durable queue for replication stream {"id": "0000000000000001", "path": "C:\\Users\\circleci\\AppData\\Local\\Temp\\TestSendWrite1627281409\\001\\replicationq\\0000000000000001"}
logger.go:130: 2022-06-23T13:00:54.457Z ERROR Error in replication stream {"replication_id": "0000000000000001", "error": "remote timeout", "retries": 1}
testing.go:1090: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestSendWrite1627281409\001\replicationq\0000000000000001\1: The process cannot access the file because it is being used by another process.
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: fix failing TestStore_BadShard on Windows
=== FAIL: tsdb TestStore_BadShard (0.09s)
logger.go:130: 2022-06-23T12:18:21.827Z INFO Using data dir {"service": "store", "path": "C:\\Users\\circleci\\AppData\\Local\\Temp\\TestStore_BadShard1363295568\\001"}
logger.go:130: 2022-06-23T12:18:21.827Z INFO Compaction settings {"service": "store", "max_concurrent_compactions": 2, "throughput_bytes_per_second": 50331648, "throughput_bytes_per_second_burst": 50331648}
logger.go:130: 2022-06-23T12:18:21.828Z INFO Open store (start) {"service": "store", "op_name": "tsdb_open", "op_event": "start"}
logger.go:130: 2022-06-23T12:18:21.828Z INFO Open store (end) {"service": "store", "op_name": "tsdb_open", "op_event": "end", "op_elapsed": "77.3µs"}
testing.go:1090: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestStore_BadShard1363295568\002\data\db0\rp0\1\index\0\L0-00000001.tsl: The process cannot access the file because it is being used by another process.
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: fix failing TestPartition_PrependLogFile_Write_Fail and TestPartition_Compact_Write_Fail on Windows
=== FAIL: tsdb/index/tsi1 TestPartition_PrependLogFile_Write_Fail/write_MANIFEST (0.06s)
testing.go:1090: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestPartition_PrependLogFile_Write_Failwrite_MANIFEST656030081\002\0\L0-00000003.tsl: The process cannot access the file because it is being used by another process.
--- FAIL: TestPartition_PrependLogFile_Write_Fail/write_MANIFEST (0.06s)
=== FAIL: tsdb/index/tsi1 TestPartition_Compact_Write_Fail/write_MANIFEST (0.08s)
testing.go:1090: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestPartition_Compact_Write_Failwrite_MANIFEST3398667527\002\0\L0-00000003.tsl: The process cannot access the file because it is being used by another process.
--- FAIL: TestPartition_Compact_Write_Fail/write_MANIFEST (0.08s)
We must close the open file descriptor otherwise the temporary file
cannot be cleaned up on Windows.
Fixes: 619eb1cae6 ("fix: restore in-memory Manifest on write error")
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: fix failing TestReplicationStartMissingQueue on Windows
=== FAIL: TestReplicationStartMissingQueue (1.60s)
logger.go:130: 2023-03-17T10:42:07.269Z DEBUG Created new durable queue for replication stream {"id": "0000000000000001", "path": "C:\\Users\\circleci\\AppData\\Local\\Temp\\TestReplicationStartMissingQueue76668607\\001\\replicationq\\0000000000000001"}
logger.go:130: 2023-03-17T10:42:07.305Z INFO Opened replication stream {"id": "0000000000000001", "path": "C:\\Users\\circleci\\AppData\\Local\\Temp\\TestReplicationStartMissingQueue76668607\\001\\replicationq\\0000000000000001"}
testing.go:1206: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestReplicationStartMissingQueue76668607\001\replicationq\0000000000000001\1: The process cannot access the file because it is being used by another process.
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: update TestWAL_DiskSize
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
* test: fix failing TestWAL_DiskSize on Windows
=== FAIL: tsdb/engine/tsm1 TestWAL_DiskSize (2.65s)
testing.go:1206: TempDir RemoveAll cleanup: remove C:\Users\circleci\AppData\Local\Temp\TestWAL_DiskSize2736073801\001\_00006.wal: The process cannot access the file because it is being used by another process.
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
---------
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
This commit fixes a data race in the WAL, which can occur when writes
and deletes are being executed concurrently. The WAL uses a buffer pool
of `[]byte` when reading the WAL. WAL entries are unmarshaled into these
buffers and passed along to the relevant methods handling the different
types of entry (write, delete etc).
In the case of deletes, the keys that need to be deleted were being
stored for later processing, however these keys were part of the backing
array of initial buffer from the pool. As such, those keys could be
written to at a future time when handling other parts of the WAL.
Re-open the last wal segment instead of creating a new one. This fixes
an issue where the last modified time of the WAL would change on
restart. It also avoids a lot of IO file churn on restart.
The default of 4096 results in writes to the WAL still requiring muliple
IOs. We had previously bumped this to 1M, but that was too high when
there are many shards. Increasing to around 16k reduces the IOs to
one or two for the workloads tested. We may want to make this
configurable in the future.
The previous sha was taken from a revision on a devel branch that I
thought would continue staying in the tree after it was merged. That
revision was rebased away and the API was changed for the logger.
This updates the usage of the logger and adds a simple package for
constructing the base logger.
The 1.0 version of zap changed the format of the default console logger
so this change moves over to this new logger instead of attempting to
retain backwards compatibility with the old format.
The fysncs due to large writes when writing to TSM files and the
WAL can eventually cause large pauses. Since we already buffer
writes, using synchronous IO reduces fsync latency by ensuring
the individiual writes hit disk. This spreads out the latecncy
across multiple writes better.
This switches all the interfaces that take string series key to
take a []byte. This eliminates many small allocations where we
convert between to two repeatedly. Eventually, this change should
propogate futher up the stack.
* introduced UnsignedValue type
* leveraged existing int64 compression algorithms (RLE, Simple 8B)
* tsm and WAL can read and write UnsignedValue
* compaction is aware of UnsignedValue
* unsigned support to model, cursors and write points
NOTE: there is no support to create unsigned points, as the line
protocol has not been modified.
There was a race in the WAL writeToLog and scheduleSync which could
lead to a writing goroutine blocking indefinitely on its syncErr channel.
The issue was that the clearing of the syncCount happenend after the
wal was unlock. If a goroutine was able to lock, write and call scheduleSync
before the existing scheduleSync goroutine returns and ran the defer to
clear the syncCount, then a new scheduleSync goroutine would not get started.
This left the writing goroutine block with nothing to signal it.
While in this state, a RLock on the engine was held. If a Lock was requested
on the engine during this time, all future writes and queries would block waiting
on the blocked wal writer.
The fix is to move the atomic clearing of syncCount before the Lock is released.
This pool was previously a pool.Bytes to avoid repetitive allocations.
It was recently switchted to a sync.Pool because pool.Bytes held onto
very larger buffers at times which were never released. sync.Pool is
showing up in allocation profiles quite frequently.
This switches the pool to a new pool that limits how many buffers are
in the pool as well as the max size of each buffer in the pool. This
provides better bounds on allocations.
The monitor goroutine ran for each shard and updated disk stats
as well as logged cardinality warnings. This goroutine has been
removed by making the disks stats more lightweight and callable
direclty from Statisics and move the logging to the tsdb.Store. The
latter allows one goroutine to handle all shards.
The current bytes.Pool will hold onto byte slices indefinitely. Large
writes can cause the pool to hold onto very large buffers over time.
Testing w/ sync/pool seems to perform similarly now so using a sync/pool
will allow these buffers to be GC'd when necessary.
Under high write load, the sync goroutine would startup, and end
very frequently. Starting a new goroutine so frequently adds a small
amount of latency which causes writes to take long and sometimes timeout.
This changes the goroutine to loop until there are no more waiters which
reduce the churn and latency.
If the sync waiters channel was full, it would block sending to the
channel while holding a the wal write lock. The sync goroutine would
then be stuck acquiring the write lock and could not drain the channel.
This increases the buffer to 1024 which would require a very high write
load to fill as well as retuns and error if the channel is full to prevent
the blocking.
Fsyncs to the WAL can cause higher IO with lots of small writes or
slower disks. This reworks the previous wal fsyncing to remove the
extra goroutine and remove the hard-coded 100ms delay. Writes to
the wal still maintain the invariant that they do not return to the
caller until the write is fsync'd.
This also adds a new config options wal-fsync-delay (default 0s)
which can be increased if a delay is desired. This is somewhat useful
for system with slower disks, but the current default works well as
is.
They rebased a revision we were previously relying upon that allowed us
to use the vanity name so we are reverting back to an older version with
the old import path.
Every write to the WAL current runs and fsync before returning. When
there are lot of concurrent writes, this can cause the WAL to bottleneck
write throughput since fsyncs are very expensive.
This changes the writeToLog to fsync on an interval to allow multiple fsyncs
calls to be batched up into one. The writeToLog behavior is the same in that
it won't return until an fsync has been performed.
Geoffrey Wossum
Phil Bracikowski
Eng Zer Jun
Sam Arnold
Daniel Moran
Yun Zhao
Tristan Su
Stuart Carnie
Mark Rushakoff
Edd Robinson
Jonathan A. Sternberg
Jason Wilder
Joe LeGasse
Jonathan A. Sternberg
Ben Johnson