shaotingcheng
shaotingcheng
For those wrong metric, the run of repair was killed and I got this log `WARN [2021-04-24 07:56:51,655] [qleartest:708d4400-a341-11eb-bfea-7ff2db79e593] i.c.s.RepairRunner - RepairRun "708d4400-a341-11eb-bfea-7ff2db79e593" does not exist. Killing RepairRunner for this...
@burmanm thanks for your reply. Indeed, I see other exceptions **HeartbeatException** in management-api log ``` ERROR [nioEventLoopGroup-3-15] 2023-05-03 09:06:31,264 NodeOpsResources.java:562 - Error when executing request com.datastax.oss.driver.api.core.connection.HeartbeatException: null at com.datastax.oss.driver.api.core.connection.HeartbeatException.copy(HeartbeatException.java:58) at...
The cause is "flush" operation takes too long. I saw some commit logs are > 1Gi ``` "May 5, 2023 @ 23:18:03.940","Completed flushing /var/lib/cassandra/data/demo1/product_data-60a74550712f11ed9db1679d9bc4cfab/nb-55917-big-Data.db (1017.550MiB) for commitlog position CommitLogPosition(segmentId=1683277730848, position=292404316)"...
I got the same timeout PT30S by executing `medusa` CLI in medusa POD ``` cassandra@cassandra-dc1-r1-sts-1:~$ medusa backup --backup-name medusa-test --mode differential [2023-05-18 09:46:07,631] WARNING: The CQL_USERNAME environment variable is deprecated...
I built my own docker image to increase the REQUEST_TIMEOUT to 120s https://github.com/k8ssandra/management-api-for-apache-cassandra/blob/master/management-api-server/src/main/java/com/datastax/mgmtapi/UnixSocketCQLAccess.java#L131 No PT30S request timeout, but sometimes Cassandra returns ERROR {code}DriverTimeoutException: Heartbeat request: timed out after 5000 ms{code}...
For those API require the dbUnixSocketFile, management-api can only handle one request at a time. Is the statement correct? @burmanm I made an example: 1. takesnapshot first ``` cassandra@cassandra-dc1-r1-sts-0:/$ date;curl...
@emerkle826 thanks for the feedback. Indeed, from the [DataStax Java-driver](https://github.com/datastax/java-driver/tree/4.x/manual/developer/common/concurrency#non-blocking), > The only place where the driver blocks is when using the synchronous API (methods declared in [SyncCqlSession](https://docs.datastax.com/en/drivers/java/4.14/com/datastax/oss/driver/api/core/cql/SyncCqlSession.html%60)), and when...
I'm not sure the reason or details, but the issue disappear after I rebooted one of the control-plane.
I can't update the livenessProbe or readinessProbe in a running k8ssandraClusrer. I thought "K8ssandraCluster.spec.cassandra.datacenters[index].containers[index].readinessProbe " can be used, but it's not. https://docs.k8ssandra.io/reference/crd/releases/k8ssandra-operator-releases/k8ssandra-operator-crds-1.5/#k8ssandraclusterspeccassandradatacentersindexcontainersindexreadinessprobe operator will show ERROR ``` 2023-05-08T02:26:22.620Z ERROR SuperuserSecretName...