| tt_ | Hello! Is ~8g of ram normal for healthmanager? running in k8 cluster 3 replicas all using about the same amount of resources. Seems a little high to me, but curious if this is expected or if not, is there anything i can do to improve resource consumption | 00:48 |
|---|---|---|
| johnsom | That seems a bit high. How are you getting the memory usage? | 01:42 |
| tt_ | Pulling it from k9s util. Will double check things via prometheus metrics next | 01:54 |
| johnsom | Mine has a resident size of ~500MB, but it only has one amphora running right now. Also, it's a 4 core VM, so the process count is low. | 02:00 |
| johnsom | If you are running it with a large number of cores, you probably have more processes spun up than you need as python will allocate one for each core on the host. You can tune this with [health_manager] health_update_threads and stats_update_threads. | 02:02 |
| johnsom | The sizing depends on the number of amphora you have and the number of controllers. | 02:02 |
| tt_ | Thank you for the info, i'll take a deeper dig into it all! | 02:18 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!