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
Currently, Tempo Stack controls how resources are distributed amongst the Tempo deployments. This might lead to the resource wasting. While some of the deployments have overcommitted limits, the other struggle to get more. Just as in my case:
You can clearly see that the Ingester does not need such high resources reserved while the Compactor practically gets OOM Killed all the time. Of course I have scaled up to a bigger instance to keep compactor alive, but then I have tons of resources being wasted.
How about Tempo Stack allowed overriding components limits? E.g like that:
Currently, Tempo Stack controls how resources are distributed amongst the Tempo deployments. This might lead to the resource wasting. While some of the deployments have overcommitted limits, the other struggle to get more. Just as in my case:
You can clearly see that the Ingester does not need such high resources reserved while the Compactor practically gets OOM Killed all the time. Of course I have scaled up to a bigger instance to keep compactor alive, but then I have tons of resources being wasted.
How about Tempo Stack allowed overriding components limits? E.g like that:
The text was updated successfully, but these errors were encountered: