Github user jinxing64 commented on the issue:
https://github.com/apache/spark/pull/18388
@tgravescs
I think it's not that hurt.
In current change, new client is compatible with the old and new shuffle
service. In our clusters, we always upgrade the client first and then server
side, which will not cause incompatible issue. The only risk here is that user
upgrades the server but still using the old client. But I find no reason they
do this. I think users usually tend to upgrade the client first and then deploy
new servers gradually.
In our cluster, there are nodemanagers failing everyday because of OOM of
shuffle service. The root cause is that shuffle service is a hot point and
there is no concurrency control.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
