65 Commits (bcdc807dd34d7a4e2e1dcfdffecd869da66e6efe)

Author	SHA1	Message	Date
Travis Patterson	4da0f63789	Remove unused Value field (#9774) ... We didn't end up needing the `Value` field because we can express everything we need to as selectors	3 years ago
Trevor Whitney	dbc3040739	Convert SeriesVolume response to prometheus-style (#9703) ... Changes the response type of the label volume stats endpoint to return volumes as prometheus-style timeseries metrics. It currently only supports instant queries, but is a necessary step to eventually supporting range queries.	3 years ago
Dylan Guedes	c0d9a2b4ba	Log push failures per-entry instead of in batches (#9720) ... **What this PR does / why we need it**: Modify how we log push errors to do it per-entry instead of accumulating all errors and then logging them all in a single pass. **Which issue(s) this PR fixes**: N/A	3 years ago
Travis Patterson	db97058a84	Series volume endpoint (#9704) ... This changes the `label_volume` endpoint to the `series_volume` endpoint. The new endpoint still returns volumes but now it does it for the requested streams defined by the selector names passed rather than individual labels. All relevant non-requested labels are aggregated into the returned results ex: Assume we have the following streams: ``` {cluster="prod", team="A", component="foo"} {cluster="prod", team="B", component="foo"} {cluster="dev", team="A", component="foo"} {cluster="dev", team="B", component="foo"} ``` - requesting `{cluster="prod"}` returns one result for all streams containing `{cluster="prod"}` - requesting `{cluster=~".+"}` returns two results for the streams containing `{cluster="prod"}` and `{cluster="dev"}` - requesting `{cluster=~".+", team=".+"}` returns four results for the streams containing: ``` {cluster="prod", team="A"} {cluster="prod", team="B"} {cluster="dev", team="A"} {cluster="dev", team="B"} ``` --------- Co-authored-by: Trevor Whitney <trevorjwhitney@gmail.com>	3 years ago
Trevor Whitney	4a56445686	Upgrade `golangci-lint` and fix linting errors (#9601) ... **What this PR does / why we need it**: Upgrade `golangci-lint` and fixes all the errors. The upgrade includes some stricter linting.	3 years ago
Travis Patterson	065bee7e72	Label Volume Endpoint (#9588) ... For a given set of matchers, returns the top N associated label/value pairs by volume. A query for `{cluster=prod}` will return ``` cluster=prod: size (total logs matching this matcher) . . . nth-label=nth-value ``` This is to service use cases where users want to understand where their log volume has come from by label without making multiple requests to the stats endpoint. Note: This PR is a monster but it's mostly plumbing. I've pointed out the most interesting bits that actually get the volumes from ingesters/indexs	3 years ago
Periklis Tsirakidis	69392de8b6	Fix duplicate label values from ingester streams (#9629)	3 years ago
Travis Patterson	6d8a184334	Revert "revert count unflused chunks (#9504)" (#9615) ... Now that the bug where huge values is fixed and known not to be this. Let's readd it. Original: https://github.com/grafana/loki/pull/9479	3 years ago
Travis Patterson	a4ef06890b	revert count unflused chunks (#9504) ... revert this commit: aeba51ab0f We seem to be dramatically overcounting bytes now which is causing oversharding This is the same query before and after: Before: ~20 GB  After: ~8 EB 	3 years ago
Travis Patterson	aeba51ab0f	Count unflushed chunks in instance stats (#9479) ... The instance wasn't counting unflushed chunks in the stats. This change checks that the chunk's `flushed time` is 0 which means it hasn't been flushed yet.	3 years ago
Peter Štibraný	90a1d4593e	Update Prometheus dependency (#9205)	3 years ago
Travis Patterson	92fca94bb5	Revert "Update Ingester Rate Calculations (#7652)" (#7991) ... This change was supposed to smooth out the per-second rate of streams so distributors could react more uniformly to streams with variable rates. In practice, it resulted in drastically more resource usage on distributors because `GetStreamRates` now responds with all streams on an ingester rather than only the streams seen in the last second. This change also didn't seem to affect sharding precision one way or another. If we'd still like to implement exponential smoothing for stream rates, it should be implemented in distributors to keep communication overhead to a minimum.	3 years ago
Travis Patterson	343e39f67b	Update Ingester Rate Calculations (#7652) ... - Use weighted moving average for rates - Keep Streams around until they are removed Previously, the rate calculator only kept the most recent second. This led to issues where bursty streams that didn't get pushes every second resulted in undersharding those streams	3 years ago
Travis Patterson	c75b822fc6	Track rates per stream and per tenant (#7564) ... It is possible for two tenants to have streams with the same labels. This PR changes the stream tracking in the `shard_tracker`, `stream_rate_calculator`, and `rate_store` to account for tenant ID when tracking streams.	3 years ago
Travis Patterson	45d59ba571	Improve rate store refresh latency (#7492) ... Rather than iterating streams and getting rates per stream, keep all necessary information in the rate calculator and just hand back the current rates when asked. This change took p99 latencies from ~1.8s to ~400ms and p50 latences from ~1s to 6ms.	3 years ago
Robert Fratto	2208ea7147	Update Prometheus dependency to latest release (v2.39.1) (#7425)	3 years ago
Travis Patterson	3e7adb2c34	Measure Per-Stream rates (#7370) ... This PR implements the ability for ingesters to measure per-stream rates at 1/s resolution. It also fixes bugs in the kind of Limits used in the RateStore and the Ingester API	3 years ago
Dylan Guedes	b1d4efab12	Loki: Per-tenant stream sharding (#7311) ... **What this PR does / why we need it**: - Move stream sharding configuration to its own package to avoid cyclic imports - Change stream sharding to be a per-tenant configuration - Change ingesters to reject whole streams due to rate-limit based on per-tenant stream sharding - Change stream sharding flags prefix from `distributor.shard-stream` to `shard-stream`	3 years ago
Travis Patterson	e3e4e27c8c	Implement ingester API to return stream rates (#7176) ... This PR introduces an API on the Ingester that reports all current streams in memory and their current rate in bytes/second. The reponse from the API contains a slice of `StreamRate` structs: ```go type StreamRate struct { StreamHash uint64 `protobuf:"varint,1,opt,name=streamHash,proto3" json:"streamHash,omitempty"` StreamHashNoShard uint64 `protobuf:"varint,2,opt,name=streamHashNoShard,proto3" json:"streamHashNoShard,omitempty"` Rate int64 `protobuf:"varint,3,opt,name=rate,proto3" json:"rate,omitempty"` } ``` In the interest of space on the caller and in transmission, only hashes and rates are sent. - `StreamHash` will be used to deduplicate replicas - `StreamHashNoShard` will be used to aggregate a sharded stream's total rate. Co-authored-by: Dylan Guedes <djmgguedes@gmail.com>	3 years ago
Travis Patterson	bc16450349	Revert "Hide duplicates caused by stream sharding (#7005)" (#7031) ... @owen-d had a suggestion that was dramatically dimpler. Revert this whole thing and the simplified PR will be forthcoming	3 years ago
Travis Patterson	a98b7ebaf4	Hide duplicates caused by stream sharding (#7005) ... As streams are sharded and resharded across different ingesters, the likelihood of duplicate logs being ingested increases dramatically. This removes duplicates at query time by dynamically removing the `__shard_stream__` label for both log and sample queries so the `stream hash` used by the `MergeEntryIterator` will match when doing comparisons for duplicates. The `stream hash` is created in either the `store` or `ingester instance` at query time so the labels are removed there. Additionally, the `MergeEntryIterator` has been modified to remove duplicates in a stream rather than just across streams. For completeness, `__shard_stream__` has also been removed from Label and Series Queries.	3 years ago
Mathew Heard	de418bf6ac	Reduce to a single code path through sendBatches regardless of limit (#6216) ... * Reduce to a single code path through sendBatches regardless of limit specified * simplify sendBatches by using -1 to specify no limit * test fix	4 years ago
Owen Diehl	4903df6ae5	Tsdb/inverted index wiring (#6252) ... * multi inverted index * delete in inverted index signature * multi inverted index * inverted index testing + skip nil indices * wires up period config aware multi-inverted-idx through ingesters * better inverted index validation * apply shipper defaults to last relevant period config, not current one. * more verbose error	4 years ago
Christian Simon	7f50a26656	Move ingester metrics around (#6275)	4 years ago
Travis Patterson	5939d5e9d8	Query filtering in the ingester and storage (#5629) ... * Add timestamp to pipeline and extractor interfaces * implement filter extractor and pipeline * implement storage filtering * implement ingester filtering * test cleanup * Refactor pipeline setup to util * linter * enable filtering based on mode * review feedback * review comments * review feedback * no need to include label filters in the test pipeline * review feedpack	4 years ago
Cyril Tovena	8f02495453	Refactoring of the Loki storage (#5833) ... * Extract index interface from the series store * moving package around * Moving more things around Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Moving more things around Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * moving package around * Moving more things around again Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Moving more things around again Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * improve composite store for more generics index store * fixes first round of tests and ensure coverage is still good * implement series method * Fixes storage tests Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Add test for GetSeries Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Make Loki compile Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * lint the whole project Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Fixes tests Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * tsdb util package usage Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Add more information of why SetChunkFilterer exists for the index Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Fixes ruler/base package Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Fixes o2 Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com>	4 years ago
Cyril Tovena	79f0e349f5	Refactor LogQL syntax into its own package (#5539) ... * Refactor LogQL syntax into its own package Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Merge upstream Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Update pkg/loghttp/series.go Co-authored-by: Susana Ferreira <ssncferreira@gmail.com> * lint Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * lint Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> Co-authored-by: Susana Ferreira <ssncferreira@gmail.com>	4 years ago
Kaviraj Kanagaraj	2db309271c	Makes `tailer.droppedStreams` slice bounded. (#5334) ... * Makes `tailer.droppedStreams` slice bounded. For slow receiver, this dropped streams may cause ingester memory to keep increasing. Sometimes OOMing the ingester. Signed-off-by: Kaviraj <kavirajkanagaraj@gmail.com> * Make droppedStreams slice size configurable Update tests Adding logging when slice is dropped Signed-off-by: Kaviraj <kavirajkanagaraj@gmail.com> * Add test for dropped stream Signed-off-by: Kaviraj <kavirajkanagaraj@gmail.com>	4 years ago
Cyril Tovena	3af7b7900e	Refactor HeapIterator into Merge and Sort Iterator. (#5281) ... * Refactor HeapIterator into merge and sort Iterator. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * lint Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com>	4 years ago
RangerCD	b26b73d5be	Remove instance level streams lock (#5179) ... * Feat(ingester): remove instance level streams lock * Fix lint * Solve Load and Delete consistency problem while pushing to ingester	4 years ago
Karsten Jeschkies	7bbed2be49	Define Ingester interface to enable wrapping. (#4988)	4 years ago
Dylan Guedes	d5783e86a3	Copy `cortex/pkg/querier/astmapper` package dependency into Loki (#4982) ... * Fork cortex `pkg/querier/asmapper` into Loki. * Move away from cortex `pkg/querier/astmapper`. * Fix lint issues. - Use _ in argument of method `pkg/querier/astmapper/subtree_folder.go` since it is not being used - Remove unnecessary cast from `subtree_folder.go`	4 years ago
Marco Pracucci	c53457feb9	Upgrade Cortex, Prometheus and Thanos (#4830) ... * Upgrade Cortex, Prometheus and Thanos Signed-off-by: Marco Pracucci <marco@pracucci.com> * Fixed imported ordering Signed-off-by: Marco Pracucci <marco@pracucci.com> * Fixed data type used for series ref when recovering from checkpoint/WAL Signed-off-by: Marco Pracucci <marco@pracucci.com> * Upgraded Thanos to fix arm build Signed-off-by: Marco Pracucci <marco@pracucci.com> * Run go mod tidy Signed-off-by: Marco Pracucci <marco@pracucci.com>	4 years ago
Karsten Jeschkies	5ece938f3c	Run all benchmarks. (#4742) ... * Run all benchmarks. * Start debugging memchunk benchmark. * Do not iterate when closed. * Pass default tenant config. * Set default limits. * Fix two more benchmarks. * Enable cron trigger again.	4 years ago
Hiroki Sakamoto	0e21f02d76	Respect shard number in series api (#4577) ... * Respect shard number in series api * Refactor codes * remove duplicate logic in instance * format with goimport * Improve complexity of getting all fingerprints	4 years ago
Owen Diehl	20515a2402	better per stream rate limits configuration options (#4236) ... * better per stream rate limits configuration options * clean up test config * pr feedback	4 years ago
Owen Diehl	668622c813	Refactor per stream rate limit (#4213) ... * uses new StreamRateLimiter * sets burst before limiter, uses requested ts * creates new non infinite rate limiters when configs change * time/rate testware * less indirection & protect rate limit access from race conditions * removes old comment * always recreate stream limiter after config change	4 years ago
Callum Styan	757eeacb88	[ingester/stream]: Add a byte stream rate limit. (#4191) ... * Add a byte stream rate limit. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Track discarded lines due to stream rate limit + use ByteSize for the config value. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Address review comments Signed-off-by: Callum Styan <callumstyan@gmail.com> * Another round of review fixes. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Fix goimports issue. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Fix duplicate flag registration + stream rate limit test. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Fix stream rate/burst limit flags yaml naming. Signed-off-by: Callum Styan <callumstyan@gmail.com> * Fix tests now that we have stream rate limit. Signed-off-by: Callum Styan <callumstyan@gmail.com> * fix some comments Signed-off-by: Callum Styan <callumstyan@gmail.com>	4 years ago
Owen Diehl	ff2ced09ea	Cleanup/unordered writes ingester config (#4192) ... * handle replaying unordered WAL into ordered configs. * ingester limiter metric, testware, correctly replayed unordered WAL into ordered config	4 years ago
Owen Diehl	066f0bf3c1	Tenant/unordered (#4130) ... * makes unordered writes a tenant config * json tag annotation	4 years ago
Owen Diehl	92a4e57448	ingester.index-shards config (#4111) ... * consistent shard mapping test & better error msg * ingester.index-shards config * loki fails to start if the ingester & schema factors arent compatible. * lint * docs * update changelog	4 years ago
Owen Diehl	572bbe15ec	Discard/ooo (#4104) ... * stream includes tenant id * expose out_of_order discarded metrics	4 years ago
Ed Welch	6e67d1ea5d	Loki/Promtail: Client Refactor (#3623) ... * move docker driver * move fluent-bit output plugin * move fluentd plugin * move logstash plugin * move promtail * match stage uses simple logql label matcher and filter expressions only * remove remaining logql imports * remove helpers package and move contents to util package * move validation and runtime out of util package * fix imports * move build and cfg packages into util * move ParseRequest function to util package * create a new package for logql models and move marshal/unmarshal code to util folder * move ErrorLabel to logqlmodel * move stats package to logqlmodel * decouple loghttp from logql by moving validation to logql package * move gcplog to tools * move PackedEntryKey constant to logqlmodel * fix stats proto import and rebuild protos * fix windows promtail tests * update drone to new dockerfile locations	5 years ago
Cyril Tovena	72dcda165a	Adds chunk filter hook for ingesters. (#3603) ... * Adds chunk filter hook for ingesters. Follow up on https://github.com/grafana/loki/pull/3569 this is the missing piece. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * That interface is not required to change. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com>	5 years ago
Ed Welch	8012362674	Loki: Per Tenant Runtime Configs (#3460) ... * Add per tenant configs mechanism * Add per tenant logging of stream creation * fix tests * enable per tenant configs on push request logging. * fixing up the stream log, changing log levels to debug	5 years ago
Cyril Tovena	55b91fc793	Fixes head chunk iterator direction. (#3383) ... * Fixes head chunk iterator direction. For backward queries, since LogQL parser we are using a heapIterator in the headchunk to re-order properly entries. But we also reverse all iterators in the memchunk code, even the headchunk which causes reversal of already reversed entries. This PR skips reversal of the headchunk. Fixes #3345 Fixes #3208 This has for side effects: - when using replication you would not dedupe data properly anymore, since the data is not correctly ordered accross batches. - when using limit it would miss entries. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Fix reversal and time filtering of headchunk. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com>	5 years ago
Owen Diehl	95052943c4	swaps mutex for atomic (#3141)	5 years ago
Owen Diehl	ed649eec37	Don't fail writes due to full WAL disk (#3136) ... * dont fail writes on full wal disk * wal full failure will cause flush on shutdown * logs the first full WAL failure	5 years ago
Cyril Tovena	cdf4a73238	Improve tailer matching by using the index. (#3090) ... * Improve tailer matching by using the index. ``` ❯ benchcmp before.txt after benchmark old ns/op new ns/op delta Benchmark_instance_addNewTailer-16 537731 1479 -99.72% benchmark old allocs new allocs delta Benchmark_instance_addNewTailer-16 244 3 -98.77% benchmark old bytes new bytes delta Benchmark_instance_addNewTailer-16 22098 121 -99.45% ``` Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com> * Review feedback. Signed-off-by: Cyril Tovena <cyril.tovena@gmail.com>	5 years ago
Owen Diehl	4d9865acd4	Adds WAL support (experimental) (#2981) ... * marshalable chunks * wal record types custom serialization * proto types for wal checkpoints * byteswith output unaffected by buffer * wal & record pool ifcs * wal record can hold entries from multiple series * entry pool * ingester uses noopWal * removes duplicate argument passing in ingester code. adds ingester config validation & derives chunk encoding. * segment writing * [WIP] wal recovery from segments * replay uses sync.Maps & preserves WAL fingerprints * in memory wal recovery * wal segment recovery * ingester metrics struct * wal replay locks streamsMtx in instances, adds checkpoint codec * ingester metrics * checkpointer * WAL checkpoint writer * checkpointwriter can write multiple checkpoints * reorgs checkpointing * wires up checkpointwriter to wal * ingester SeriesIter impl * wires up ingesterRecoverer to consume checkpoints * generic recovery fn * generic recovery fn * recover from both wal types * cleans up old tmp checkpoints & allows aborting in flight checkpoints * wires up wal checkpointing * more granular wal logging * fixes off by 1 wal truncation & removes double logging * adds userID to wal records correctly * wire chunk encoding tests * more granular wal metrics * checkpoint encoding test * ignores debug bins * segment replay ignores out of orders * fixes bug between WAL reading []byte validity and proto unmarshalling refs * conf validations, removes comments * flush on shutdown config * POST /ingester/shutdown * renames flush on shutdown * wal & checkpoint use same segment size * writes entries to wal regardless of tailers * makes wal checkpoing duration default to 5m * recovery metrics * encodes headchunks separately for wal purposes * merge upstream * linting * addresses pr feedback uses entry pool in stream push/tailer removes unnecessary pool interaction checkpointbytes comment fillchunk helper, record resetting in tests via pool redundant comment defers wg done in recovery s/num/count/ checkpoint wal uses a logger encodeWithTypeHeader now creates its own []byte removes pool from decodeEntries wal stop can error * prevent shared access bug with tailers and entry pool * removes stream push entry pool optimization	5 years ago