You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Per #5507 (comment), now that both #5507 and #5508 have merged, I'm revisiting the mgbench queries with CSUP to make sure they all still run ok and see if their performance has improved. When doing this, I've found that the bench1/q4 query now panics where it didn't previously.
Test data is s3://brim-sampledata/mgbench/bench1.csup. The older CSUP that used to run ok via super dev vector query pre-#5507 has been renamed to s3://brim-sampledata/mgbench/bench1-pre-super-5507.csup.
$ super -version
Version: v1.18.0-180-g1f0c0cea
$ SUPER_VAM=1 super -c '
from "bench1.csup"
| where
machine_group=="Servers"
and cpu_wio > float32(0.99)
and log_time >= 2016-12-01T00:00:00Z
and log_time < 2017-01-01T00:00:00Z
| summarize
spikes := count()
by machine_name
| sort -r spikes
| head 10'
panic: runtime error: index out of range [0] with length 0
goroutine 2486 [running]:
github.com/brimdata/super/runtime/vam/op/summarize.(*countByString).countDict(0xc000b94200, 0xdb65d3c?, {0x0, 0x0, 0x1?}, 0x0)
/Users/phil/work/super/runtime/vam/op/summarize/agg.go:204 +0x1a5
github.com/brimdata/super/runtime/vam/op/summarize.(*countByString).update(0xf4f9020?, {0xc000b941c0?, 0x0?, 0xc000a03c08?}, {0xc000b941d0?, 0x574792f8?, 0xc000b95f80?})
/Users/phil/work/super/runtime/vam/op/summarize/agg.go:151 +0x134
github.com/brimdata/super/runtime/vam/op/summarize.(*Summarize).consume(0xc0006c6d10, {0xc000b941c0, 0x1, 0x2}, {0xc000b941d0, 0x1, 0x1})
/Users/phil/work/super/runtime/vam/op/summarize/summarize.go:102 +0x4cc
github.com/brimdata/super/runtime/vam/op/summarize.(*Summarize).Pull.func1({0xc000b941c0, 0x2, 0xc000a03d80?})
/Users/phil/work/super/runtime/vam/op/summarize/summarize.go:83 +0x78
github.com/brimdata/super/vector.Apply(0x10?, 0xf78fa78?, {0xc000b941c0?, 0x2?, 0xf58c620?})
/Users/phil/work/super/vector/apply.go:17 +0xa3
github.com/brimdata/super/runtime/vam/op/summarize.(*Summarize).Pull(0xc0006c6d10, 0x80?)
/Users/phil/work/super/runtime/vam/op/summarize/summarize.go:82 +0x13b
github.com/brimdata/super/runtime/vam/op.(*combineParent).run(0xc0011f0540)
/Users/phil/work/super/runtime/vam/op/combine.go:110 +0x3f
created by github.com/brimdata/super/runtime/vam/op.(*Combine).Pull.func1 in goroutine 2481
/Users/phil/work/super/runtime/vam/op/combine.go:42 +0x35
The text was updated successfully, but these errors were encountered:
Repro is with super commit 1f0c0ce.
Per #5507 (comment), now that both #5507 and #5508 have merged, I'm revisiting the mgbench queries with CSUP to make sure they all still run ok and see if their performance has improved. When doing this, I've found that the bench1/q4 query now panics where it didn't previously.
Test data is
s3://brim-sampledata/mgbench/bench1.csup
. The older CSUP that used to run ok viasuper dev vector query
pre-#5507 has been renamed tos3://brim-sampledata/mgbench/bench1-pre-super-5507.csup
.The text was updated successfully, but these errors were encountered: