Zach Leslie

I spent a little time with this last week, let me see if I can summarize. In doing a little research, I ran into a couple documents which describe what...

https://github.com/grafana/dskit/pull/185 could use another test pass, but I doubt I'll have the time in the next month. Specifically, I want to know more about the DNS handling when results come...

Related is https://github.com/grafana/tempo/issues/1544, and the configuration I had to tune to get IPv6 to work properly in Tempo https://github.com/zalegrala/tempo/blob/inet6/operations/jsonnet/microservices/common.libsonnet#L17 I'm `zleslie` in the community slack if you need a hand...

I was able to get the test to fail after many iterations using this snippet ``` while go test -v ./tempodb/... -run '(TestTimeWindowBlockSelectorBlocksToCompact)' -count=1; do echo "trying again";done ```

Have we seen this recently?

Thanks for filing an issue. I'll see if I can reproduce this based on your config.

I've been able to reproduce all but the third scenario there. In my case testing with a 75Mib trace, I would get some connection timeouts but not an OOM. For...

Thanks for the information. In my test, I did query the frontend directly. In my case, port `3200` is mapped to the query-frontend, so just issuing `curl -s http://localhost:3200/api/traces/4fe154b38dee96fb19c30b4e2091e559` was...

This was enough to see the memory spike within the pod: `while true; do ps -o rss,vsz; sleep .01; done`

Just confirming with my team, but I don't think there would be an impact here with the number of queriers. Each querier will require a little more memory on the...