Suns storage servers again are better here and it stands to reason. not really curated into a collection of graphs of analyzed raw data to help solve an actual bottleneck or other issue. sure they provide pretty graphs (some of them) but its usually rather simplistic raw data. I think that none of the current front ends really power use dtrace. and show a pre and post view of what is pending before execute is pressed.ģ> monitoring. My suggestions then for any GUI front end to ZFS is thatġ> any ACTION features of the GUI should show the actual terminal commands that the pending operation will accomplish and be editable so the admin can tweek it if necessary.Ģ> clearly warn when a pending activity will result in a irreversible data structure change or result in data destruction. and MANY things in ZFS are un-reversable are require the full destruction and rebuild at best or DATA LOSS at the other endĢ> GUI administration tends to lower the bar and stagnates or eliminates the learning curve that I feel anyone managing a ZFS array needs to learn so graphically it all looks good but the action fails to do what the admin THOUGHT was going to happen. One glaring issue with them all is that administration of ZFS from a GUI only interface at some point almost always leads an admin to do something stupid.ġ> because its not always evident WHAT or HOW something is actually going to be done. ![]() ![]() Thus far I have dabbled with many front ends, Sun's zfs storage server (one of the better ones), freenas, napp-it etc and very timely now that joyent have abandoned solars as their head for zfs development and moved linux to the head spot moving forward as the lead operating system
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |