Commit graph

58 commits

Author SHA1 Message Date
Fabian Reinartz d226411a2a db: delete old blocks during reload
Windows requires blocks to be closed before deleting their directories.
This adds a set of deleteable blocks to reload(), which then deletes
them without causing disruption to querying.
2017-11-03 20:41:24 +01:00
Fabian Reinartz 6a10761b50 Remove GetTabWriter from tsdb package 2017-10-19 18:14:37 +02:00
Fabian Reinartz 91a154d228 Fix block printing in cmd/main 2017-10-11 11:02:57 +02:00
Bas Harenslak a41dad30fb Move printing db.blocks to main.go 2017-10-02 22:48:47 +02:00
Bas Harenslak 9945a67bff Replace columnize by stdlib tabwriter 2017-10-02 22:29:51 +02:00
Bas Harenslak e190c7c78d Add list blocks command to CLI 2017-10-01 22:18:50 +02:00
Fabian Reinartz 162a48e4f2 Create series with ID recorded in WAL when reading it back 2017-09-19 11:31:16 +02:00
Fabian Reinartz 99d39174f6 Merge branch 'master' of github.com:prometheus/tsdb 2017-09-18 11:20:45 +02:00
Fabian Reinartz 8214dc82a7 Remove infinite block in benchmark 2017-09-18 11:20:25 +02:00
Fabian Reinartz e45bb1d328 Merge pull request #142 from krasi-georgiev/107-swap-cobra-with-kingpin
replaced cobra with kingpin for the tsdb cli tool
2017-09-12 19:28:35 +02:00
Krasi Georgiev 92d0414993 replaced cobra with kingpin
Signed-off-by: Krasi Georgiev <krasi.root@gmail.com>
2017-09-10 12:35:02 +03:00
Fabian Reinartz 6892fc6dcb Finish old WAL segment async, default to no fsync
We were still fsyncing while holding the write lock when we cut a new
segment. Given we cannot do anything but logging errors, we might just
as well complete segments asynchronously.

There's not realistic use case where one would fsync after every WAL
entry, thus make the default of a flush interval of 0 to never fsync
which is a much more likely use case.
2017-09-08 18:41:12 +02:00
Fabian Reinartz c36d574290 Replace single head lock with granular locks
This adds various new locks to replace the single big lock on
the head. All parts now must be COW as they may be held by clients
after initial retrieval.
Series by ID and hashes are now held in a stripe lock to reduce
contention and total holding time during GC. This should reduce
starvation of readers.
2017-09-05 14:41:39 +02:00
Fabian Reinartz 3951d8cc29 Fix selection of compactable blocks 2017-08-04 12:02:34 +02:00
Fabian Reinartz 9c4235532e Fix compaction selection after creating new heads
This fixes the case where between block creations no compaction
plans are ran. We were not compacting anything in these
cases since the on creation the most recent head block always had
a high timestamp of 0.
2017-07-13 16:15:13 +02:00
Goutham Veeramachaneni 9a2ab732b5 Choose ranges by retention and min-block-duration
Signed-off-by: Goutham Veeramachaneni <cs14btech11014@iith.ac.in>
2017-07-12 17:30:34 +02:00
Fabian Reinartz 05e411a8eb Improve heuristic to spread chunks across block 2017-06-08 11:30:32 +02:00
Fabian Reinartz 39df7e2bba Switch blocks to ULID directories, drop sequenc numbers 2017-05-18 16:09:30 +02:00
Fabian Reinartz 285bc07030 Switch append refs to string 2017-05-18 10:56:57 +02:00
Fabian Reinartz c8438cfc81 Add mutex profiling to benchmark 2017-05-14 11:51:56 +02:00
Fabian Reinartz 778103b450 Add liecence file and headers 2017-04-10 20:59:45 +02:00
Fabian Reinartz 10c7c9acbe Adjust import names to new repository organisation 2017-04-04 11:27:26 +02:00
Goutham Veeramachaneni 761e4768f3
Lint and Vet Fixes 2017-03-19 21:35:01 +05:30
Fabian Reinartz 87805fb83f Remove Partitioned* code 2017-03-06 17:34:49 +01:00
Fabian Reinartz cc0a7c8279 Create alloc and inuse space heap profile 2017-03-02 21:54:30 +01:00
Fabian Reinartz ffb24a98f4 Add missing unlock, run debug endpoint in benchmark 2017-03-02 21:54:02 +01:00
Fabian Reinartz db5c88ea9a Misc compaction fixes 2017-02-28 08:40:51 +01:00
Fabian Reinartz b281e4e39b Accept prometheus.Registerer in constructor 2017-02-28 07:26:24 +01:00
Fabian Reinartz a3b47c4929 Create default logger for DB 2017-02-19 16:04:37 +01:00
Fabian Reinartz 9c7a88223e Add full encode/decode WAL cycle test 2017-02-14 21:55:50 -08:00
Fabian Reinartz 79944a5912 Break out WAL into segment files 2017-02-13 23:53:19 -08:00
Fabian Reinartz 9c76624df2 Add initial retention cutoff 2017-02-09 17:54:26 -08:00
Fabian Reinartz 012cf4ef25 Count writer references on head blocks 2017-02-04 11:53:52 +01:00
Fabian Reinartz 5a1c8eaa0e Fix missing appends after reference lookups 2017-02-02 11:10:17 +01:00
Fabian Reinartz 30efe4a58c Support writing to multiple head blocks
This is an initial (and hacky) first pass on allowing
appending to multiple blocks simultaniously to avoid
dropping samples right after cutting a new head block.
It's also required for cases like the PGW, where a scrape may
contain varying timestamps.
2017-02-01 15:57:28 +01:00
Fabian Reinartz c20cc44b06 Add docs, write sequence number to meta.json 2017-01-29 08:11:47 +01:00
Fabian Reinartz 5fb01d41aa Use new Prometheus text format parser 2017-01-16 21:29:53 +01:00
Fabian Reinartz fde69dab49 Use buffer pool for head appenders 2017-01-12 20:03:44 +01:00
Fabian Reinartz a317f252b9 Expose series references to clients
This exposes a reference number of a series represented by a label set
to clients.
Subsequent samples can be directly added via the reference rather than
repeatedly passing in the full labels. This drasitcally speeds up the
append process.

The appender chain uses different sections of the reference number for
assignment to child appenders and invalidating reference numbers as
necessary.

Clients can either pass out reference numbers themselves or have their
own optimized lookup, i.e. by directly associating unparsed metric
descriptors strings with reference numbers.
2017-01-12 20:00:54 +01:00
Fabian Reinartz 80affd98a8 Add barrier to benchmark writer
This adds a barrier to avoid issues with unfair goroutine scheduling
that causes some fake scrapers to run away from the other ones.
2017-01-11 13:01:30 +01:00
Fabian Reinartz c32a94d409 Unexport HeadBlock, export Block interface 2017-01-10 15:41:57 +01:00
Fabian Reinartz d86e8a63c7 Report correct number of appended samples 2017-01-10 11:17:37 +01:00
Fabian Reinartz 300f4e2abf Use separate lock for series creation
This uses the head block's own lock to only lock if new series were
encountered.
In the general append case we just need to hold a
2017-01-06 18:10:50 +01:00
Fabian Reinartz c61b310210 Naive size-based compaction
This adds naive compaction that tries to compact three
blocks of roughly equal size.
It decides based on samples present in a block and has no
safety measures considering the actual file size.
2017-01-06 13:53:05 +01:00
Fabian Reinartz 3ed2c2a14b Rename Partition to regular DB, DB to PartitionedDB 2017-01-06 11:40:09 +01:00
Fabian Reinartz cabd7e4ebc Add small testdata file 2016-12-22 19:15:37 +01:00
Fabian Reinartz ede733ab6c Extract labels package 2016-12-21 09:39:01 +01:00
Fabian Reinartz ee217adc7e Redfine append interface, remove old Prometheus storage from bench 2016-12-21 00:02:37 +01:00
Fabian Reinartz e561c91d53 Implement proper buffered iterator
This adds a proper duration based lookback buffer for series iterators
to allow advancing sequentially while remaining able to calculate time
aggregating functions such as `rate` backwards.

It uses an array ring buffer to minimize heap allocations for
potentially hundreds of thousands of series for a single query.
2016-12-14 21:14:44 +01:00
Fabian Reinartz 3a528c3078 Write plain postings list index 2016-12-10 09:44:00 +01:00