devmadhuu commented on code in PR #6272:
URL: https://github.com/apache/ozone/pull/6272#discussion_r1559327763
##########
hadoop-ozone/recon/src/main/java/org/apache/hadoop/ozone/recon/scm/ReconStorageContainerManagerFacade.java:
##########
@@ -383,32 +420,41 @@ private OzoneConfiguration getReconScmConfiguration(
* Start the Recon SCM subsystems.
*/
@Override
- public void start() {
+ public void start() throws IOException {
if (LOG.isInfoEnabled()) {
LOG.info(buildRpcServerStartMessage(
"Recon ScmDatanodeProtocol RPC server",
getDatanodeProtocolServer().getDatanodeRpcAddress()));
}
+ reconSCMSyncScheduler = Executors.newScheduledThreadPool(1, threadFactory);
+ registerSCMDBTasks();
+ try {
+ scmMetadataManager.start(ozoneConfiguration);
+ } catch (IOException ioEx) {
+ LOG.error("Error staring Recon SCM Metadata Manager.", ioEx);
+ }
+ reconTaskController.start();
+ long initialDelay =
ozoneConfiguration.getTimeDuration(OZONE_RECON_SCM_SNAPSHOT_TASK_INITIAL_DELAY,
+ OZONE_RECON_SCM_SNAPSHOT_TASK_INITIAL_DELAY_DEFAULT,
TimeUnit.MILLISECONDS);
+
+ // This schedules a periodic task to sync Recon's copy of SCM metadata
+ // with SCM metadata rocks DB. Gets full snapshot or delta updates.
Review Comment:
I do not want to remove this task because this task queries pipelines and
their state from SCM which are not yet flushed to SCM DB and I still see ratis
transaction threshold as 10000 , crossing which it forces to flush in DB. Also
I am not sure why but see this task is also having a call for syncing
operational state on dead nodes.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
