[
https://issues.apache.org/jira/browse/YARN-8060?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16428626#comment-16428626
]
Shane Kumpf commented on YARN-8060:
-----------------------------------
Thanks for the patch, [~billie.rinaldi]! This is a much needed check given the
delay we see with the IP address becoming available. I have tested the IP and
DNS portions of this readiness check and am getting the desired results.
Few suggestions:
# There is one case where I think having a default readiness check that
depends on IP might be an issue, which is when {{--net=none}}. In that case the
container will never get an IP address. While {{--net=none}} is the only case I
can come up with, there may be others where this check could be problematic.
Could we consider a configuration that would allow for disabling the default
check?
# With this patch, the container correctly stays in a RUNNING_BUT_UNREADY
state until the default readiness check passes, but as a user it's unclear why
the container is still in that state. Could we add logging to the AM that shows
the status of the readiness checks?
# {{ServiceRegistryUtils.registryDNSLookupExists}} could use additional
comments. Can you elaborate on the need for the second lookup?
# The service API docs were updated, but the description for the HTTP
readiness check doesn't mention that the DEFAULT checks will also be executed.
I'd like to see the DEFAULT check mentioned somewhere in the docs.
# The checkstyle issues look valid if you could address those.
> Create default readiness check for service components
> -----------------------------------------------------
>
> Key: YARN-8060
> URL: https://issues.apache.org/jira/browse/YARN-8060
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: yarn-native-services
> Reporter: Billie Rinaldi
> Assignee: Billie Rinaldi
> Priority: Major
> Attachments: YARN-8060.1.patch
>
>
> It is currently possible for a component instance to have READY status before
> the AM retrieves an IP for the container. We should make sure the IP has been
> retrieved before marking the instance as READY.
> This default probe could also have an option to check for a DNS entry for the
> instance's hostname if a DNS address is provided.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
