[
https://issues.apache.org/jira/browse/HDDS-13890?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ivan Andika updated HDDS-13890:
-------------------------------
Description:
Currently if we would like to migrate the SCM from (scm1, scm2, scm3) to (scm4,
scm5, scm6), all the datanodes need to be restarted 2 times with updated
ozone.scm.nodes.<service> # Before migration: Update ozone.scm.nodes to (scm1,
scm2, scm3, scm4, scm5, scm6)
# After (scm1, scm2, scm3) are decommissioned: Update ozone.scm.nodes to
(scm4, scm5, scm6)
As mentioned in HDDS-12391, rolling restarting all the datanodes might take a
while. For large datanodes fleet this might take a lot of time (days or even
weeks).
It might be good to support dynamic reconfiguration of SCM endpoints in DN to
prevent restarts. A possible flow
* Admin update the "ozone.scm.nodes" to the new value (with some new nodes and
removed nodes)
* DN will compare the new and previous configuration and find the SCM
endpoints to add and remove
* DN will add the SCM endpoints (e.g. SCMConnectionManager#addSCMServer) and
then remove the SCM endpoints (e.g. SCMConnectionManager#removeSCMServer).
was:
Currently if we would like to migrate the SCM from (scm1, scm2, scm3) to (scm4,
scm5, scm6), all the datanodes need to be restarted 2 times with updated
ozone.scm.nodes.<service> # Before migration: Update ozone.scm.nodes to (scm1,
scm2, scm3, scm4, scm5, scm6)
# After (scm1, scm2, scm3) are decommissioned: Update ozone.scm.nodes to
(scm4, scm5, scm6)
As mentioned in HDDS-12391, rolling restarting all the datanodes might take a
while. For large datanodes fleet this might take a lot of time (days or event
weeks).
It might be good to support dynamic reconfiguration of SCM endpoints in DN to
prevent restarts. A possible flow
* Admin update the "ozone.scm.nodes" to the new value (with some new nodes and
removed nodes)
* DN will compare the new and previous configuration and find the SCM
endpoints to add and remove
* DN will add the SCM endpoints (e.g. SCMConnectionManager#addSCMServer) and
then remove the SCM endpoints (e.g. SCMConnectionManager#removeSCMServer).
> Datanode supports dynamic configuration of SCM
> ----------------------------------------------
>
> Key: HDDS-13890
> URL: https://issues.apache.org/jira/browse/HDDS-13890
> Project: Apache Ozone
> Issue Type: Task
> Reporter: Ivan Andika
> Assignee: Ivan Andika
> Priority: Major
>
> Currently if we would like to migrate the SCM from (scm1, scm2, scm3) to
> (scm4, scm5, scm6), all the datanodes need to be restarted 2 times with
> updated
> ozone.scm.nodes.<service> # Before migration: Update ozone.scm.nodes to
> (scm1, scm2, scm3, scm4, scm5, scm6)
> # After (scm1, scm2, scm3) are decommissioned: Update ozone.scm.nodes to
> (scm4, scm5, scm6)
> As mentioned in HDDS-12391, rolling restarting all the datanodes might take a
> while. For large datanodes fleet this might take a lot of time (days or even
> weeks).
> It might be good to support dynamic reconfiguration of SCM endpoints in DN to
> prevent restarts. A possible flow
> * Admin update the "ozone.scm.nodes" to the new value (with some new nodes
> and removed nodes)
> * DN will compare the new and previous configuration and find the SCM
> endpoints to add and remove
> * DN will add the SCM endpoints (e.g. SCMConnectionManager#addSCMServer) and
> then remove the SCM endpoints (e.g. SCMConnectionManager#removeSCMServer).
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
