Nothing obvious from the logs you posted.
Generally speaking, replaying commit log is often the culprit when a
node takes a long time to start. I have seen many nodes with large
memtable and commit log size limit spending over half an hour replaying
the commit log. I usually do a "nodetool flush" before shutting down the
node to help speed up the start time if the shutdown was planned. There
isn't much you can do about unexpected shutdown, such as server crashes.
When that happens, the only reasonable thing to do is wait for the
commit log replay to finish. You should see log entries related to
replaying commit logs if this is the case.
However, if you don't find any logs related to replaying commit logs,
the cause may be completely different.
On 19/01/2022 11:54, Paul Chandler wrote:
Hi all,
We have upgraded a couple of clusters from 3.11.6, now we are having issues
when we restart the nodes.
The node will either hang or take 10-30 minute to restart, these are the last
messages we have in the system.log:
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,267 FileUtils.java:545 -
Deleting file during startup:
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb-184-big-Summary.db
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,268 LogTransaction.java:240 -
Unfinished transaction log, deleting
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb-185-big-Data.db
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,268 FileUtils.java:545 -
Deleting file during startup:
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb-185-big-Summary.db
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,269 LogTransaction.java:240 -
Unfinished transaction log, deleting
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb-186-big-Data.db
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,270 FileUtils.java:545 -
Deleting file during startup:
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb-186-big-Summary.db
INFO [NonPeriodicTasks:1] 2022-01-19 10:08:23,272 LogTransaction.java:240 -
Unfinished transaction log, deleting
/var/lib/cassandra/data/system/table_estimates-176c39cdb93d33a5a2188eb06a56f66e/nb_txn_unknowncompactiontype_bc501d00-790f-11ec-9f80-85
8854746758.log
INFO [MemtableFlushWriter:2] 2022-01-19 10:08:23,289 LogTransaction.java:240
- Unfinished transaction log, deleting
/var/lib/cassandra/data/system/local-7ad54392bcdd35a684174e047860b377/nb_txn_flush_bc52dc20-790f-11ec-9f80-858854746758.log
The debug log has messages from DiskBoundaryManager.java at the same time, then
it just has the following messages:||
DEBUG [ScheduledTasks:1] 2022-01-19 10:28:09,430 SSLFactory.java:354 -
Checking whether certificates have been updated []
DEBUG [ScheduledTasks:1] 2022-01-19 10:38:09,431 SSLFactory.java:354 -
Checking whether certificates have been updated []
DEBUG [ScheduledTasks:1] 2022-01-19 10:48:09,431 SSLFactory.java:354 -
Checking whether certificates have been updated []
DEBUG [ScheduledTasks:1] 2022-01-19 10:58:09,431 SSLFactory.java:354 -
Checking whether certificates have been updated []
It seems to get worse after each restart, and then it gets to the state where
it just hangs, then the only thing to do is to re bootstrap the node.
Once I had re bootstrapped all the nodes in the cluster, I thought the cluster
was stable, but I have now got the case where the one of the nodes is hanging
again.
Does anyone have an ideas what is causing the problems ?
Thanks
Paul Chandler