Bringing together the Apache Cassandra experts from the community and DataStax.

Want to learn? Have a question? Want to share your expertise? You are in the right place!

Not sure where to begin? Getting Started

 

question

opaseka_191284 avatar image
opaseka_191284 asked Erick Ramirez commented

What does "Flushing of largest memtable, not done" mean?

Hi,

We have a cluster 16CPU/128Gb 500Gb SSD - data disk, 500Gb SSD - solr.data disk and 50Gb SSD - commit logs. Cluster with 3 DSE Search nodes. Xmx=32Gb. DSE Cassandra 6.8.10. RF=3

After some time (2-8 hours), all cluster nodes stop responding. We see a ton of messages in debug.log:

ColumnFamilyStore.java:1692 - Flushing of largest memtable, not done, max live ratio 0.32 less than min ratio 0.33

Can someone explain me what is can be and where I need to look to find root cause?

Thanks in advance.

dse
10 |1000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 1.0 MiB each and 10.0 MiB total.

1 Answer

Erick Ramirez avatar image
Erick Ramirez answered Erick Ramirez commented

That log entry just indicates that a memtable was not flushed to disk because the data it contains is less than a third of the memtable segment. On its own, it is not indicative of any problem which is why the messages are logged at DEBUG level instead of WARN or ERROR.

More importantly, it doesn't have any relation at all to your nodes no longer responding. You need to review the logs around the time a node becomes unresponsive for clues.

If you need help with diagnosing problems with your DSE cluster, please log a ticket with DataStax Support so one of our engineers can assist you directly. Cheers!

1 comment Share
10 |1000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 1.0 MiB each and 10.0 MiB total.

Erick Ramirez avatar image Erick Ramirez ♦♦ ·

I've noted that one of your colleagues logged a ticket with DataStax Support. Cheers!

0 Likes 0 ·