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
In order to avoid having servers that have been created at the same time to get coalesced at the exact same time, we could just initiate the sequence for this server with its srvid. Then if you create 3 servers at the same time, they coalesce will be offset by one cycle...
The text was updated successfully, but these errors were encountered:
I don't know how frequent your pattern is but there is clearly need for that so I agree it would be nice to do.
That being said, I don't think that simply using the srvid id would work that great in practice for your use case. The whole point of the coalesce is to benefit from compression so clearly the first dozen or few dozen of servers will do a coalesce and use resources for no disk usage gain. How about introducing a new min_coalesce or something like that and initiate the sequence to a value between that minimal value and coalesce - 1? This way it would still help avoiding a storm a coalesce (which I think was @marco44 original problem) but also still mitigate the disk usage problem you have. That combined with a connection pooler in front of the collector's repository connection and things could be quite good.
In order to avoid having servers that have been created at the same time to get coalesced at the exact same time, we could just initiate the sequence for this server with its srvid. Then if you create 3 servers at the same time, they coalesce will be offset by one cycle...
The text was updated successfully, but these errors were encountered: