lvisterin commented on pull request #1524:
URL: https://github.com/apache/zookeeper/pull/1524#issuecomment-723870998
> I saw you originally guarded the address recreation with the following
condition:
>
> ```
> if (address.isUnresolved()) { ... }
> ```
>
> Why did you decide to remove this condition? Anyway, this code shouldn't
run frequently, so I'm fine to recreate the address every time... but I'm
curious: having the `if (address.isUnresolved())` didn't solve the problem?
In the container environment I am running in, it is possible to get the old
address of a certain Zookeeper instance when it is rescheduled. In such case
the address will be "resolved", just to the wrong one, and it will again fail
in a loop.
That said I think it is reasonable to say Zookeeper does not need to support
this scenario and that my DNS just has to be fixed. Most users won't run into
this. So let me know if you'd prefer for me to add it back, because I do have
other workarounds for this error case.
> sure, go ahead, I'll approve the change and merge it. Let us know when you
think you finished
Have not gotten to committing this yet because I am dealing with some other
strange things happening in our setup and need to make sure this patch is good.
I'll leave a comment when I know more.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
