deepakbnsl
deepakbnsl
"can hit this with spdk. " - any workaround to avoid this situation or mitigation from openEBS side?
> restart the io-engine pod proactively after creating a number of snapshots so that blobstore metadata is synced to disk. Instead of io-engine restart, Is there any other mechanism (api...
> io-engine process crashes for some reason(like a SIGKILL) what can be the unplanned causes for SIGKILL? And how to ensure a graceful shutdown for planned terminations (say host restart)?
> 32MB cluster size Is there any side effect of this?
> And guess what, it also solves https://github.com/spdk/spdk/issues/3609 problem (17TiB mkfs hang) that would be super awesome!!
> The cluster size is not tunable via helm chart today though. We'll see how to make it configurable. any timeline?