Thanks for pinging on IRC. The server appears unresponsive, but it's actually load-shedding. It is serving requests, just a large number of long-cpu requests are coming in right now, so beyond a certain number, they won't get serviced.

I'll install a slightly updated version of the package that should respond a little better here. It could also be helpful to have a mod_qos type thing set up at the frontend proxies to limit too many concurrent requests from the outside. Investigating other caching/config tweaks also.

did the following:
- installed a new upstream release of elfutils debuginfod (0.187) built for f34
- bumped up some caching parameters in /etc/sysconfig/debuginfod

Will keep monitoring a while to see if this is enough. If so, will update the VM gradually to F36, which has the newer elfutils build, and enshrine the new cache configs in ansible.

Things looking stable for the last hour+.

Thanks @fche

I guess lets close this now?