influx_inspect walks the data and wal directories building a list of
files to export. It then opens, reads, and exports each. If the file was
deleted between the time it was added to the list and the time the
inspect tool attempts to read it, the file is now skipped without
emitting an error.
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.
Measurement name and field were converted between []byte and string
repetively causing lots of garbage. This switches the code to use
[]byte in the write path.
Benchmark improvements with this change:
benchmark old ns/op new ns/op delta
BenchmarkExportTSMFloats_100s_250vps-4 23206480 10279106 -55.71%
BenchmarkExportTSMInts_100s_250vps-4 17995000 5762310 -67.98%
BenchmarkExportTSMBools_100s_250vps-4 17067605 4235467 -75.18%
BenchmarkExportTSMStrings_100s_250vps-4 54846997 34682568 -36.76%
BenchmarkExportWALFloats_100s_250vps-4 23459937 10436297 -55.51%
BenchmarkExportWALInts_100s_250vps-4 18747150 6236062 -66.74%
BenchmarkExportWALBools_100s_250vps-4 17988273 4814358 -73.24%
BenchmarkExportWALStrings_100s_250vps-4 59700802 35815739 -40.01%
benchmark old allocs new allocs delta
BenchmarkExportTSMFloats_100s_250vps-4 201442 51738 -74.32%
BenchmarkExportTSMInts_100s_250vps-4 201442 51728 -74.32%
BenchmarkExportTSMBools_100s_250vps-4 201441 51638 -74.37%
BenchmarkExportTSMStrings_100s_250vps-4 404092 201584 -50.11%
BenchmarkExportWALFloats_100s_250vps-4 250322 75627 -69.79%
BenchmarkExportWALInts_100s_250vps-4 250323 75617 -69.79%
BenchmarkExportWALBools_100s_250vps-4 250321 75527 -69.83%
BenchmarkExportWALStrings_100s_250vps-4 452868 225291 -50.25%
benchmark old bytes new bytes delta
BenchmarkExportTSMFloats_100s_250vps-4 5170539 2351789 -54.52%
BenchmarkExportTSMInts_100s_250vps-4 5143189 2331276 -54.67%
BenchmarkExportTSMBools_100s_250vps-4 3724951 2143780 -42.45%
BenchmarkExportTSMStrings_100s_250vps-4 17131400 10796281 -36.98%
BenchmarkExportWALFloats_100s_250vps-4 4487868 1468109 -67.29%
BenchmarkExportWALInts_100s_250vps-4 4458395 1452359 -67.42%
BenchmarkExportWALBools_100s_250vps-4 2838719 1258755 -55.66%
BenchmarkExportWALStrings_100s_250vps-4 16787201 10010700 -40.37%
Also, after improving those benchmarks, I did a time-filtered export on
a 450MB TSM file to a 21GB plain text output, with and without the
bufio.BufferedWriter.
Without buffering, it took about 263s, and with buffering, it took about
60s, for a delta of about -77%.
Mark Rushakoff
Stuart Carnie
Jonathan A. Sternberg
David Norton
Jason Wilder
joelegasse
Cory LaNou