51 Commits (main)

Author	SHA1	Message	Date
Ashwanth	54f9723af1	chore(xcap): adds exporter to summarise capture as a structured log line (#20099)	3 weeks ago
Ashwanth	728b308edf	chore: update read stats to use xcap (#20095)	3 weeks ago
benclive	2ce207fbdb	chore: Use shared zstd encoders during dataobj build (#20044)	4 weeks ago
Robert Fratto	806bcd5dd7	chore(dataobj): stop preallocating memory for page builders (#20043)	1 month ago
benclive	a6de577b3a	chore: Fix page skipping with empty predicates (#19602)	2 months ago
Ashwanth	bfd70602dd	chore: correct dataobj reader rows read return value (#19583)	2 months ago
Robert Fratto	7556192423	chore(dataobj): use single Zstd decoder (#19508)	2 months ago
benclive	2dac5ef8a7	chore: Write empty message values in logs section writer (#19359)	3 months ago
Sophie Waldman	af6983e9de	fix: Add special case handling for comparing empty and nil values (#19348) ... We can have a predicate checking for an empty string, e.g. {"foo" = ""}. If a row has no value for the column "foo", it will return a nil value for that column. Thus, when we go to compare the row with the predicate, we will be comparing an empty value to a nil value. This change adjusts the comparison check so that we correctly handle this special case.	3 months ago
Christian Haudum	a3017f2b13	chore(dataobj-consumer): Sort logs object-wide (#19231) ... Sorting logs globally (object-wide per tenant) removes overlapping time ranges of sections in the objects. Sections contain logs from multiple streams, and the ingest lag of streams may vary. This means that although logs of individual sections are sorted by timestamp, the overall sorting of logs is not guaranteed. Therefore sections are sorted with a k-way merge (SortMerge), which does over-query data in case the query would reach its result limit early. Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	3 months ago
Christian Haudum	2ecf07d52c	chore(dataobj): Fix panic in column builder (#19153) ... There is an edge case in the column builder where it panics when the AppendNulls() call tries to append more NULL values than the specified MaxPageRows setting when backfilling rows. In that case the builder would panic because canAppend would always return false even when the previous page was flushed. Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	4 months ago
Christian Haudum	4c81ebc7c3	chore(dataobj): Add option to set target row limit for pages (#19128) ... This PR introduces an option for the column builder to limit the number of rows in a page. It can be set for logs object and index objects separately using `-dataobj-consumer.max-page-rows` and `-dataobj-index-builder.max-page-rows` respectively. Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	4 months ago
Robert Fratto	f6091a67d1	chore(engine): move to toggleable section prefetching (#19142) ... dataset.readerDownloader was originally introduced in #16429, an attempt to balance peak memory usage of reading a section with read times by downloading a configurable size of pages in advance. In practice, each roundtrip to object storage adds too much of a latency hit, and we've started to set the cache limit high enough to ensure that each reader only needs a single prefetch. Given what we've found, it no longer makes sense to control peak memory usage via the prefetch size. Other options, such as downloading directly to disk, may be explored in the future. In the meantime, this PR removes the ability to specify a cache size. All non-pruned pages will be bulk requested using the range reader (#19067) on the first read call. Pages which have left the potential read window will continue to be eagerly removed for garbage collection. However, we don't want to prefetch when the dataset is entirely in memory, which is the case when the logs section builder is performing k-way merge over in-memory sections. To lower the memory usage of builders, prefetching is configurable. For this initial PR, prefetching is only disabled for the logs section builder; all other reads force prefetching. Signed-off-by: Robert Fratto <robertfratto@gmail.com>	4 months ago
Robert Fratto	515ebea4f2	chore(dataobj): track total download time of page metadata/data (#18896)	4 months ago
Robert Fratto	595715146f	chore(dataobj)!: introduce new centralized dataset encoding format (#18871) ... Signed-off-by: Robert Fratto <robertfratto@gmail.com>	4 months ago
Ashwanth	5203a0c9ad	chore(dataobj): add reader stats (#18694)	5 months ago
Robert Fratto	a5c722cbed	chore(dataobj): list pages across all columns in one request (#18678)	5 months ago
Robert Fratto	9129d8cdd0	chore(engine): use columnar logs.Reader for DataObjScan (#18577)	5 months ago
benclive	077e4196ef	chore: Add indexes to LogQL benchmarks (#18500)	5 months ago
benclive	1f0edcd4ac	feat: Initial index-builder implementation (#18297)	6 months ago
Robert Fratto	26cb50e65a	chore(dataobj): Improve performance of dataset.Value API (#18339) ... Co-authored-by: Ashwanth <iamashwanth@gmail.com>	6 months ago
benclive	a238816751	perf: Use map for InPredicate when reading dataobj (#18325)	6 months ago
Ashwanth	5c02493fe2	chore(dataobj): avoid reading a predicate column more than once (#18324)	6 months ago
Robert Fratto	c5ac922e55	chore(dataobj): Use pooled buffers when decoding pages (#17955)	7 months ago
Christian Haudum	48f1d0ebc0	chore(dataobj): Decompress zstd compressed pages upfront (#17841) ... This PR changes the decompression of data object pages, so that they are decompressed fully upfront into a bytes buffer, instead of passing a streaming zstd decompressor (reader) to the page decoder. This this allows to release the zstd decompressor to the pool earlier.Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	7 months ago
Ashwanth	3ad1a64bc1	chore(dataobj): fixes a memory aliasing bug in dataobj reader (#17537)	8 months ago
Ashwanth	5961923836	chore(dataobj): apply predicates sequentially (#17283)	8 months ago
Ashwanth	42e2407137	chore(dataobj): backfill sparse rows using EncodeN (#16485)	8 months ago
Ashwanth	7dd4ab4b09	chore(dataobj): correct stats calculation for prepredicate bytes read (#17279)	8 months ago
benclive	9e68fb01f4	feat(dataobj): Add query stats collection to the dataobj readers (#17128)	9 months ago
benclive	e5784d7a82	perf: Reuse Values in dataobj.Reader to reduce allocs (#16988) ... Co-authored-by: Robert Fratto <robertfratto@gmail.com>	9 months ago
Ashwanth	42eed6ded5	chore(dataobj): introduce InPredicate to replace long chain of OR operators (#16919)	9 months ago
Robert Fratto	9be95a3e8c	chore(dataobj): reduce memory overhead of merging logs stripes (#16833)	9 months ago
benclive	979f7614bc	perf: Optimization pass for reading from dataobjs (#16747)	9 months ago
Robert Fratto	cd11367148	chore(dataobj): Wire dataset.Reader into dataobj.LogsReader and dataobj.StreamsReader (#16701) ... Co-authored-by: Ashwanth Goli <iamashwanth@gmail.com>	10 months ago
Robert Fratto	abf3772366	perf(dataobj): improve parallelized read spead of Zstd (#16632)	10 months ago
Paul Rogers	27431b7e7e	chore: Linting update for new golangci (#16572)	10 months ago
Robert Fratto	7f0dc01e21	chore(dataobj): introduce more efficient dataset reading (#16429) ... This commit introduces a new API, dataset.Reader, which scans through values in datasets more efficiently than dataset.Iter. As a result, dataset.Iter has been removed in favour of the new Reader. dataset.Reader is more efficient than dataset.Iter because: - It avoids using iter.Pull; the number of values flowing through coroutines added a significant amount of CPU overhead. - It supports efficient uses of predicates, where non-predicate columns are only loaded after finding matching rows from predicate columns; this avoids loading unnecessary pages. - It adds initial support for bulk page downloading, so that we need fewer roundtrips to object storage when iterating over a dataset. (This is not fully implemented yet; see below). - Early performance benchmarks see a ~3.4x speed improvement over reading 7GB of uncompressed logs, going from 192 MB/s (with dataset.Iter) to 652 MB/s (with dataset.Reader). This commit is truly massive, and to avoid it getting any larger, I'm leaving some follow up work to additional commits: - Updating dataobj.LogsReader and dataobj.StreamsReader to use the new API. - Adding support for downloading pages in batches; dataset.Reader will request multiple pages at once, but our decoder still unwraps that into several GET requests to object storage. Signed-off-by: Robert Fratto <robertfratto@gmail.com> Co-authored-by: Cyril Tovena <cyril.tovena@gmail.com>	10 months ago
Christian Haudum	f6fcc1194e	chore(dataobj): Decode values in batches (#16418) ... Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	10 months ago
Owen Diehl	66889ec51d	feat(dataobj): cardinality estimation (#16233)	11 months ago
Robert Fratto	66b2721cfa	chore(dataobj): do not panic on unrecognized compression type (#16088)	11 months ago
Robert Fratto	5ccc6790cb	chore(dataobj): support computation of min/max values in pages and columns (#16015)	11 months ago
Christian Haudum	ed863fe1be	chore: Fix Loki arm builds (#15936) ... Signed-off-by: Christian Haudum <christian.haudum@gmail.com>	11 months ago
Robert Fratto	948f5c5f3e	chore(dataobj): Reintroduce sorting of the logs section (#15906)	11 months ago
Robert Fratto	127ad5326f	chore(dataobj): support for building and flushing data objects (#15780)	12 months ago
Robert Fratto	5980396b76	chore(dataobj): fix bug with partial reads of strings (#15767)	12 months ago
Robert Fratto	aec8e968da	chore(dataobj): track non-NULL values in pages/columns (#15758)	12 months ago
Robert Fratto	c581218c4f	chore(dataobj): add Dataset with iteration and sorting support (#15652)	12 months ago
Robert Fratto	f2fc0c26c2	chore(dataobj): column building (#15634) ... This commit adds the ability to accumulate sequences of dataset.Value into a column, which is split up across multiple pages. Each page is broken down into two parts: * A bitmap-encoded sequence of booleans, where 1 indicates a row has a value and 0 indicates the row is NULL, and * the encoded sequence of non-NULL values, whose encoding is determined by the column options. The sequence of non-NULL values is then optionally compressed. This commit also includes initial support for reading these columns, starting with internal-only helper utilities for unit tests.	12 months ago
Robert Fratto	6042d0712e	chore(dataobj): add bitmap encoding (#15629)	12 months ago