![]() If you want a true successor of OpenSolaris due the desktop and a server edition, OpenIndiana that is based on pure ongoing Illumos is it. If this is a storage server only, I would prefer OmniOS due the stable and long term stable editions (optionally with paid commercial support). Without a paid subscription you are not able to get these updates what means that an original Solaris 11.4 from two years back has more and more compatibility problems. As Solaris is like Windows 10 now frozen on version 11.4 with rolling updates (ex 11.4.SRU24), you have no chance of a new free version every 2 years. While Oracle Solaris with native ZFS was still the fastest ZFS server in my tests, the important ZFS features are now also in Illumos and Open-ZFS like ZFS encryption, sequential/sorted resilvering, removeable vdev, SMB 3.1.1 (ZFS/kernelbased SMB server) and more.įor private use to must face an additional problem. As only pool version 28/5 is compatible between Oracle Solaris with native ZFS and Open-ZFS you must backup your data (or move to a pool v28/5), then import the pool v28/5 or create a new pool and restore data to switch to the free Solaris forks. The distributions based on the free Solaris fork Illumos (NexentaStor, OpenIndiana, OmniOS, SmartOS etc) are all on pool v5000 with features. This was my first Solaris install and I have learned a lot along the way which has been fine but I have been putting this off for a long time and now the time has come. ![]() Once I have the pool sorted out I am going to do a reinstall of Solaris 11.4 and clean it all up. I can't get Smartmontools to install because I am on Solaris 11.3 and it does the dummy spit about the certificate when I try to install it. Once I do the entire pool I will expand it. ![]() This wasn't the first time it had done that and given they are about eight years old I decided it was time to bite the bullet and got some replacements that are larger anyway. I was only using iostat as an inventory tool. I would have done it if necessary because better safe than sorry. ![]() Good to know that I don't have to scrub because the resilver is about eight hours to complete. Unfortunately I only have 6 bays and they are all full which is why I had to go this way instead of adding an extra one and then removing the old one. While doing the identification I made a list as you suggest in there for next time. Current napp-it gives a smart warning on some critical raw smart values.Ĭlick to expand.Yes, these new disks have the WWN on them but the old ones didn't. ![]() Also Smartmontools can give informations about a needed or suggested replace. Real problems that require a replace are reported by ZFS itself (too many errors, offline, checksum errors). Iostat is more an inventory of disks and events since last reboot. After the replacement is finished, remove the old disk (many disk controller support hot add/remove).Īn additional scrub is not needed as a resilver has read all data already. This does not affect redundancy level like it would happen when you remove a disk to replace in slot. If you have enough bays, add a new disk additionally and start a disk replace old > new. If you haven't done, power off and create the list first to avoid wrong replacements. When you use napp-it you can create and printout a graphical disk map of your storage case. A list of disks (WWN, a factory disk number) to disk bays is always recommended, either you write it down in a list (WWN is usually printed on the disk). ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |