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
Setting a zipkin collector and then unsetting it with set cluster setting trace.zipkin.collector=''; currently deadlocks the database. All goroutines end up stuck in zipkin-go-opentracing.(*HTTPCollector).Collect+0x42.
This also reveals a gap in our testing which we should plug once that fix goes in.
The text was updated successfully, but these errors were encountered:
nvanbenschoten
added
C-bug
Code not up to spec/doc, specs & docs deemed correct. Solution expected to change code/behavior.
A-tracing
Relating to tracing in CockroachDB.
labels
Sep 24, 2018
33287: vendor: update thrift and zipkin-go-opentracing r=nvanbenschoten a=nvanbenschoten
Fixes#30576.
Informs #30774.
openzipkin-contrib/zipkin-go-opentracing@4c9fbcb...f0f479aapache/thrift@327ebb6...2b7365c
I audited every commit in the zipkin-go-opentracing diff and every
commit to the `lib/go` package of the thrift diff.
Release note (bug fix): Updated Zipkin library to avoid deadlock when
stopping Zipkin tracing.
Co-authored-by: Nathan VanBenschoten <nvanbenschoten@gmail.com>
Setting a zipkin collector and then unsetting it with
set cluster setting trace.zipkin.collector='';
currently deadlocks the database. All goroutines end up stuck inzipkin-go-opentracing.(*HTTPCollector).Collect+0x42
.I have a patch out to fix this here: openzipkin-contrib/zipkin-go-opentracing#113. We should make sure that patch goes in before 2.1 and should backport it to 2.0.
This also reveals a gap in our testing which we should plug once that fix goes in.
The text was updated successfully, but these errors were encountered: