Files
049323c40f Add Parquet block format (#1479)
* starting parquet support

starting parquet support. some things are still todo:  compaction, find trace by ID

Co-Authored-By: Annanay Agarwal <[email protected]>
Co-Authored-By: Martin Disibio <[email protected]>

* add parquet reusable streaming block

* Add parquet compaction

* Add findTraceByID logic and parquet -> proto conversion

Signed-off-by: Annanay <[email protected]>

* Switch findtracebyid to readrow until we get SeekToRow working

* Fix error handling in bufferedReaderAt. Add various instrumentation and log messages

* Use jsonpb to roundtrip event attributes

* Improve buffered reader when reading near end of file. More instrumentation and tidiness

* Add totalBlockBytes instrumentation. Fix inspectedBytes instrumentation

* parquet.FindTraceByID check bloom

* Make compator flush on trace count until we get the row group size target working

* Support maxDuration in parquet search

* Correct attribute translation for array/kvlist

Signed-off-by: Annanay <[email protected]>

* extend roundtrip test

Signed-off-by: Annanay <[email protected]>

* Use default page buffer size

* Fix size url param to be uint64

* Increase parquet searches to 1MB stripes

* Add test for findtracebyid

* make row group size configurable

Signed-off-by: Annanay <[email protected]>

* Record parquet footer size in block meta and short-circuit header/footer reads

* Add tests for SubstringPredicate

* Switch compaction to compact blocks of same version, use re-usable buffer in findtracebyid

Signed-off-by: Annanay <[email protected]>

* Add search support for remaining well-known attributes and span status, add search tests

* Benchmark substringPredicate

* Make block version configurable, default to v2

* memoize substringPredicate

* Update segmentio/parquet-go to catch SeekToRow fix, revert hack in FindTraceByID

Signed-off-by: Annanay <[email protected]>

* Add multi block prefetch iterator for use in compactor

Signed-off-by: Annanay <[email protected]>

* Add BufferedWriter with outgoing queue to flush traces async

Signed-off-by: Annanay <[email protected]>

Make compactor block complete async

Signed-off-by: Annanay <[email protected]>

* First pass CLI to dump schema and column sizes from block

Signed-off-by: Annanay <[email protected]>

* honor search start/end time range

* Improve error message, add blockID

Signed-off-by: Annanay <[email protected]>

* Fix race in InstrumentedPredicate with atomics

* Make bufferedReaderAt not single-threaded

* Debug SeekToRow panic, add test data from vulture tenant

Signed-off-by: Annanay <[email protected]>

* fix test

* repair tests

* Address println warnings

* Experiment with bookmark prefetch

Signed-off-by: Annanay <[email protected]>

Remove quitCh from bookmark

Signed-off-by: Annanay <[email protected]>

* fix prefetch on the bookmarks

Signed-off-by: Annanay <[email protected]>

* lint

* tracker -> RowNumber, lint

* delete unused predicates

* lint

* lint

* mod vendor

* remove pool from buffered writer queue

Signed-off-by: Annanay <[email protected]>

* post merge cleanup

Signed-off-by: Annanay <[email protected]>

* Close quitCh instead of attempting to signal to it

Signed-off-by: Annanay <[email protected]>

* split multiblock and prefetch iterators, refactor around interface

Signed-off-by: Annanay <[email protected]>

* Split search into 2 phases to avoid loading traceid/display columns unless there is a match

* Make columnIterator.SeekTo do something

* update to catch segmentio/parquet-go#220

Signed-off-by: Annanay <[email protected]>

* Fix column index to store entire 32-byte trace ID string. lint

* Make parquet honor the adjusted start/end times that come from the wal

* post merge cleanup: use compaction metric callbacks

Signed-off-by: Annanay <[email protected]>

* serverless sum/mods

Signed-off-by: Annanay <[email protected]>

* parquetquery predicates handle column index not present

* Make parquet compaction really honor input block start/end times, add universal test at tempodb-level

* parquet skip page indexes while searching to reduce latency on openfile

* More testing around iterators and combiner

Signed-off-by: Annanay <[email protected]>

* Update to latest parquet lib to catch rle bitpack fix

* Review feedback:  LatestEncoding -> DefaultEncoding

* Pass block footerSize through serverless

* mod vendor for serverless

* Fix tests

* add blockID to error message in block iterator

Signed-off-by: Annanay <[email protected]>

* increase readvalues buffer, add more span instrumentation

Signed-off-by: Annanay <[email protected]>

* changelog and docs

* Remove the use of async writer in vparquet compactor

Signed-off-by: Annanay <[email protected]>

Co-authored-by: Annanay Agarwal <[email protected]>
2022-06-23 16:00:18 +05:30

46 lines
1.1 KiB
Go

package brotli
import (
"encoding/binary"
"math/bits"
"runtime"
)
/* Copyright 2010 Google Inc. All Rights Reserved.
Distributed under MIT license.
See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
*/
/* Function to find maximal matching prefixes of strings. */
func findMatchLengthWithLimit(s1 []byte, s2 []byte, limit uint) uint {
var matched uint = 0
_, _ = s1[limit-1], s2[limit-1] // bounds check
switch runtime.GOARCH {
case "amd64":
// Compare 8 bytes at at time.
for matched+8 <= limit {
w1 := binary.LittleEndian.Uint64(s1[matched:])
w2 := binary.LittleEndian.Uint64(s2[matched:])
if w1 != w2 {
return matched + uint(bits.TrailingZeros64(w1^w2)>>3)
}
matched += 8
}
case "386":
// Compare 4 bytes at at time.
for matched+4 <= limit {
w1 := binary.LittleEndian.Uint32(s1[matched:])
w2 := binary.LittleEndian.Uint32(s2[matched:])
if w1 != w2 {
return matched + uint(bits.TrailingZeros32(w1^w2)>>3)
}
matched += 4
}
}
for matched < limit && s1[matched] == s2[matched] {
matched++
}
return matched
}