Reloader
Reloader copied to clipboard
Simulteous Reloads maxing out cluster resources
Is there any way to not have the reloader try to do all the reloading simultaneously, and, instead work it's way through the list with different concurrency?
Currently, there is no such way. If there are concurrent events happening, the reloader will try to resolve those events concurrently as well. If you are running reloader in cluster scope, then you can try to run reloader on namespace scope where it's needed to mitigate the issue.
@nerney nice suggestion!
Possible options:
- reloader per namespace
- add support in reloader to run multiple instances in HA mode
- add support of some sort of queuing
any other thoughts?
Marking it as stale and closing it. Please reopen if required.