#general
- a
ancient-stone-50795
09/21/2015, 6:58 PM@enough-analyst-54434: I’m fine with a release. We are running acceptance tests on 0.0.48 right now.\ I don’t think your bug affects us. (since we don’t use go) - b
bored-art-40741
09/21/2015, 6:58 PMIt is not, @witty-crayon-22786 - b
bored-art-40741
09/21/2015, 6:58 PMNot even close - w
witty-crayon-22786
09/21/2015, 6:58 PMso i don't think threading issues are the first place i'd look - e
enough-analyst-54434
09/21/2015, 6:58 PMCool - thanks folks - w
witty-crayon-22786
09/21/2015, 6:58 PM@bored-art-40741: hm, when i hit the cache, all of my cores are saturated - w
witty-crayon-22786
09/21/2015, 6:59 PMsomething post hit might be slower though. - b
bored-art-40741
09/21/2015, 6:59 PMThat's for decompress, yeah - w
witty-crayon-22786
09/21/2015, 6:59 PMer, no: all the way through "the dot part" of the cache hit - w
witty-crayon-22786
09/21/2015, 6:59 PMpost dots, obviously not saturated - b
bored-art-40741
09/21/2015, 6:59 PMWe start slowing down "during" the dots - w
witty-crayon-22786
09/21/2015, 6:59 PMhm. - b
bored-art-40741
09/21/2015, 7:00 PMAnd we're talking ~500 dots or so - b
bored-art-40741
09/21/2015, 7:00 PMI'll run another repro, keep an eye on htop, and report back - w
witty-crayon-22786
09/21/2015, 7:03 PM@bored-art-40741: oh... i just remembered where i've seen saturation. when the artifacts already exist locally - w
witty-crayon-22786
09/21/2015, 7:03 PMwhen they're remote i'm network bound - b
bored-art-40741
09/21/2015, 7:03 PMAh - b
bored-art-40741
09/21/2015, 7:03 PMI'm testing remote here - b
bored-art-40741
09/21/2015, 7:03 PMThough I can test both easily - w
witty-crayon-22786
09/21/2015, 7:03 PMtry local to take that out of the equation. - b
bored-art-40741
09/21/2015, 7:04 PMgood call - h
happy-kitchen-89482
09/21/2015, 7:04 PMSo how are artifacts different now? Do they include the jar as well as the loose classfiles or something? - b
bored-art-40741
09/21/2015, 7:04 PMYes - w
witty-crayon-22786
09/21/2015, 7:05 PMyea... the jar is in there because i was lazy. so they're a bit larger. - h
happy-kitchen-89482
09/21/2015, 7:05 PMwell, that seems like a prime culprit then. - b
bored-art-40741
09/21/2015, 7:05 PMWhy? - w
witty-crayon-22786
09/21/2015, 7:05 PMit's a hypothesis, yea - h
happy-kitchen-89482
09/21/2015, 7:05 PMdecompression time? - b
bored-art-40741
09/21/2015, 7:05 PMDecompression happens in workers - w
witty-crayon-22786
09/21/2015, 7:05 PMjust one big file though... should be the best case