New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Increase grpc retry to 1000 in xdock #1419
Increase grpc retry to 1000 in xdock #1419
Conversation
Signed-off-by: Pavol Loffay <ploffay@redhat.com>
* Make grpc reporter default and add retry Signed-off-by: Pavol Loffay <ploffay@redhat.com> * Polish Signed-off-by: Pavol Loffay <ploffay@redhat.com> * Fix port Signed-off-by: Pavol Loffay <ploffay@redhat.com> * Polish Signed-off-by: Pavol Loffay <ploffay@redhat.com> * Use higher retry Signed-off-by: Pavol Loffay <ploffay@redhat.com> * Increase retry to 100 Signed-off-by: Pavol Loffay <ploffay@redhat.com>
Codecov Report
@@ Coverage Diff @@
## master #1419 +/- ##
======================================
Coverage 100% 100%
======================================
Files 164 164
Lines 7502 7502
======================================
Hits 7502 7502 Continue to review full report at Codecov.
|
Short term solution, as if still fails, it would take a long time to report - but useful to see if this makes things more stable. |
Yep this is our goal. The current configuration waits 1000*50ms with 10% jitter so around 5s -+10%. |
Restarting the build in jaegertracing/jaeger-client-java#602 |
The build still failed. |
This is unfortunate. We could go even higher to 10000. Isn't there problem with the PR itself. It failed on
|
It's possible, I didn't have time to investigate. I just know that PRs in many client libs have been failing recently on crossdock. |
I got the reason. I remembered that once I had trouble with nginx which caches the name resolve result forever by default, so I wrote a workaround.
I wrote:
On one of my environments that tends to fail (my other environments tend to succeed) could build successfully for several times. |
timeout this time... 😞 https://travis-ci.org/jaegertracing/jaeger-client-java/jobs/505333040 |
@iori-yja what are you referring to? |
@yurishkuro This commit is intended to work around the behavior and a little redundant. If the same problem happens in other PRs in client libs, possibly fix it in jaeger repository should be better (adding wait_for in crossdock's docker-compose, for example). |
I think the Java PR may be at fault for the failure itself, e.g. a Go client PR was successful (jaegertracing/jaeger-client-go#379). |
OK, it's figured out. The fault is in my change. |
Related to d8ab694#r32679636
Signed-off-by: Pavol Loffay ploffay@redhat.com