Hi
Pretty new to TrueNAS, wondering if anyone is using their install as storage backing for an ESXi datastore and if so what is your experiences with sparse zvols. I've tried a couple of installs using 12.0 u8.1, one 4 mirrored vdevs forming an all SSD pool and the other a single mirrored vdev with spinning rust. They both seem to exhibit the same issues:
1. The disk usage reported by TrueNAS is not reported back to ESXi, so all that compression goodness counts for nothing as the datstore thinks its full when it isnt e.g. pool size 2.4TB, used according to VMware 2.1TB, used according to TrueNAS 650GiB.
2. I'm getting disk warnings in VMare kernel logs - "Space utilization on thin-provisioned device naa.6589cfc0000007de18bfaca99a20ec45 exceeded configured threshold. Affected datastores (if any): xxx - according to the VMware blurb this is coming from the storage array. But I don't see any error messages or otherwise in any TrueNAS log. These appear to be generated even when ESXI thinks the datastore is below 80% full, never mind truenas; TrueNAS will state something like 28% full. Note these arent the VMware generated alerts.
3. Manual unmap commands appear to fail - "Max no space retries (10) exceeded for caller Fil3_SetFileLength (status 'No space left on device')" - however I do see some VAAI delete stats using esxitop.
Is there something I am missing or not understanding properly?
Thank you in advance for any guidance.
Pretty new to TrueNAS, wondering if anyone is using their install as storage backing for an ESXi datastore and if so what is your experiences with sparse zvols. I've tried a couple of installs using 12.0 u8.1, one 4 mirrored vdevs forming an all SSD pool and the other a single mirrored vdev with spinning rust. They both seem to exhibit the same issues:
1. The disk usage reported by TrueNAS is not reported back to ESXi, so all that compression goodness counts for nothing as the datstore thinks its full when it isnt e.g. pool size 2.4TB, used according to VMware 2.1TB, used according to TrueNAS 650GiB.
2. I'm getting disk warnings in VMare kernel logs - "Space utilization on thin-provisioned device naa.6589cfc0000007de18bfaca99a20ec45 exceeded configured threshold. Affected datastores (if any): xxx - according to the VMware blurb this is coming from the storage array. But I don't see any error messages or otherwise in any TrueNAS log. These appear to be generated even when ESXI thinks the datastore is below 80% full, never mind truenas; TrueNAS will state something like 28% full. Note these arent the VMware generated alerts.
3. Manual unmap commands appear to fail - "Max no space retries (10) exceeded for caller Fil3_SetFileLength (status 'No space left on device')" - however I do see some VAAI delete stats using esxitop.
Is there something I am missing or not understanding properly?
Thank you in advance for any guidance.
