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
After upgrading to 6.3.1 the memory usage is higher than with previous versions.
Also, fragmentation rises to 1.3 from the usual 1.05 with previous versions, the workload is the same
after enabling defrag, memory rises really fast and fragmentation rises compared to its tween server without active defrag
To reproduce
I don't have a clue on how to reproduce it, our ussage is the following
Each shard of our cluster has about 470M keys, with a long key like ( r:1567:{d2}0c5731-d778-4b2b-9a57-3daf9926fd8e ) and stores only a 4-digit integer
MEMORY USAGE r:1567:{d2}0c5731-d778-4b2b-9a57-3daf9926fd8e
(integer) 104
About 40k read qps
About 7% of the keys are deleted each day, and about 60% of the keys are updated every day
Expected behavior
Similar memory consumption and fragmentation compared to previous versions
Additional information
Any additional information that is relevant to the problem.