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
{{ message }}
This repository has been archived by the owner on Aug 23, 2023. It is now read-only.
For kafka, cassandra based on our own experience we can make the following suggestions
Kafka: minimum 3 nodes. 8core 32GB ram, 1TB SSD disk. With 3-way replication, each node can comfortably handle up to 60k metrics/second.
Cassandra: minimum 3 nodes. 8core 64GB ram, up to 4TB SSD disk.
Storeage needs vary greatly based on retention policy. Assuming a retention of 10s:8d,1m:60d,30m:2y
and 3-way replication each node can support up to 600k active series.
So a 3 node Kafa cluster + a 3 node cassandra cluster could support 1.8million series sent a 10s interval.
For metrictank resources spefically our current numbers are out of date and we now have a much larger pool of instances to analyse to get improved numbers.
@woodsaj is most familiar with this
The text was updated successfully, but these errors were encountered: