Hot answers tagged zfs
I would work with a ZFS professional or vendor who specializes in ZFS-based solutions. You're talking about 100TB of data, and at that scale, there's too much opportunity to screw this up. ZFS is not an easy thing to get right; especially when you incorporate high-availability and design for resilience. I wouldn't plan on half-filling storage enclosures or ...
If you have a pool built from two disks as a raid-0, your data was striped over both disks. You cannot recover any data from a raid-0 when one disk is missing. Yout only option is a professional data rescue service that can repair the faulted disk.
Solaris 11 ZFS uses NFSv4 ACLs rather than POSIX ACLs. Linux doesn't have NFSv4 ACLs on ZFS at all, and seems like it never will. While NFSv4 ACLs are a superset of POSIX ACLs, seems like only Solaris can transfer/translate POSIX ACLs to NFSv4 ones during file moving/copying. So, concluding, I don't see a way to preserve ACL in ZFS snapshots. Either use ...
First of all, installing FreeBSD on a Sun hardware is a serious crime, you could be decapitated in some countries (however, I think that several years in prison camp on Chukotka will be enough) ! Second, 10.x version family includes a bsdinstall pretty capable of installing FreeBSD on ZFS root without the need of emulating sunrise by hand. Third, there's ...
Why are you passing the disks through a RAID Controller? JBOD would make more sense when using ZFS. You could run into Problems because of your controller. Anyway, its save to just detach and re-attach the disk. You could also try to replace the disk (without really replacing it: zpool replace pool disk) Let it resilver and scrub again.
ZFS stripes data across top-level devices (vdevs or block targets). If you expect to be able to tolerate the loss of a top-level devices (in your case a single block target), then ZFS does not meet your needs.
I don't recommend this type of passthrough setup anymore. It's a lose-lose on most counts, especially with reliability and performance. It can certainly be done, but the safest solution (especially with a hypervisor like ESXi) is to just use a supported RAID controller and local storage. If you want ZFS storage, build a standalone ZFS storage system.
I am not sure I would use ZFS on Linux for such a setup, as ZoL remain a somewhat "moving target". Regarding your RAID card, if it can be configured in JBOD, there is no problem. However, if it only work in RAID mode, I would change it for a JBOD/HBA adapter. Anyway, as suggested by ewwhite, I would ask to a professional ZFS verdor/consultant.
I'm going to include the answer to this question for FreeBSD for sake of completeness. According to man du: -A Display the apparent size instead of the disk usage. This can be helpful when operating on compressed volumes or sparse files.
Replying to your notes: network utilization should not be the culprit, as standard Ethernet frames are 1500 bytes large at most (excluding jumbo frames). The reduced read performance seems related to something wrong with FS and device alignment. Can you double check this? Anyway, it can be also due to more fragmentation smaller blocks are less cache ...
Only top voted, non community-wiki answers of a minimum length are eligible