Title
o
Orrin
12/09/2022, 5:51 AMI had a weird error today “query aborted [fd=5408]
Any ideas?
a
Andrey Pechkurov
12/09/2022, 6:58 AMHi Orrin,
What's the full error message? It should be "timeout, query aborted [fd=5408]" or "remote disconnected, query aborted [fd=5408]"
o
Orrin
12/10/2022, 7:58 PMToday it was “[-1] timeout, query aborted [fd=5028]
a
Andrey Pechkurov
12/12/2022, 7:35 AMHi Orrin,
You may want to increase
query.timeout.secOr perhaps you could try re-writing your queries, so that they work faster.
o
Orrin
12/12/2022, 2:52 PMso what's weird about this is that I do have to run some massive queries so my conf file already has this: query.timeout.sec=1200
moreover, the actual query only takes ~600ms first run, ~200ms subsequent.
I'm guessing the default is higher than 1 second anyway so even if conf file is somehow being ignored, something seems off?
what's particularly weird is that if i submit the query via the web interface, it runs fine (how I get the timings)
i know the query is the same as I use the regurgitated version; {"query":"select distinct ExpirationDate from PJ where RICRoot = 'SPX.U';", error:"[-1] timeout, query aborted [fd=4396]", position:0}
but, if I reboot the db, it starts working again.
a
Andrey Pechkurov
12/12/2022, 3:13 PMquery.timeout.sec=1200that's a big value
are you observing the error on pgwire?
o
Orrin
12/12/2022, 11:47 PMno it's on the REST api.
a
Andrey Pechkurov
12/13/2022, 7:13 AMCan you reproduce this issue reliably?
o
Orrin
12/13/2022, 6:12 PMi thought you might ask me this. will log a bug if i can get any replicable context. thx.
a
Andrey Pechkurov
12/13/2022, 6:18 PMYeah, it's very hard to get rid of such bugs without a reproducer 🥲
o
Orrin
12/13/2022, 8:33 PMlogged an issue. https://github.com/questdb/questdb/issues/2865
logs might be more helpful than i thought at first glance.
a
Andrey Pechkurov
12/14/2022, 7:37 AMMany thanks!
The bugfix PR is on the way: https://github.com/questdb/questdb/pull/2866