landS Posted January 28, 2015 Share Posted January 28, 2015 Good day crew - I am using 2 1 TB WD Red drives as a cache pool (v6b12) After booting up TOWER2 the 2 Cache drives show as 2 TB worth of unprotected Cache available. By navigating to MAIN/CACHE/CACHE and clicking onto Balance (-dconvert=raid1 -mconvert=raid1) I can convert the array to a Raid 1 and I see 1 TB worth of data post Balance. How do I keep the balance / raid 1 selection persistent? Thanks everyone FYI I can fully saturate my networks writes to a raid 1 cache pool... NO need for additional speed of a raid 0 . Link to comment
BRiT Posted February 1, 2015 Share Posted February 1, 2015 Is there a problem? I think he's asking how does he keep this working after a reboot/power cycle. I assume he tried a reboot and the RAID-1 BTRFS Cache Pool he setup isn't persisted, so he'd have to redo the same steps again to make it work again. Link to comment
garycase Posted February 1, 2015 Share Posted February 1, 2015 Clearly that's what he's asking. landS => I infer from your question that when you reboot the array the cache is reset to RAID-0 ... is that correct? I don't have a cache, so I can't confirm the settings -- but if you've made and applied these changes they certainly should be persistent [be sure you're checking any "Apply" button that may be on the configuration page]. If the setting is not retained after a shutdown/reboot, I'd send Tom a note [ [email protected] ] and outline the issue ... you could just include a link to this thread. Link to comment
landS Posted February 1, 2015 Author Share Posted February 1, 2015 thanks Gary - when we have power back up consistently (middle of a blizzard here) I will pound away on this. The first 2 reboots after starting up a cache pool I had to rebalance for the capacity shown on the cache drive to indicate a raid-1. Link to comment
landS Posted February 1, 2015 Author Share Posted February 1, 2015 strange - this may have been caused from a disabled disk in the array. ran a new confg, removed the bad data disk (which had no data on it), and now it is not replicating. Link to comment
garycase Posted February 2, 2015 Share Posted February 2, 2015 Good ... glad it's working okay now. I suspect this "glitch" was related to the power issues you had ... hopefully it will behave now Link to comment
landS Posted February 2, 2015 Author Share Posted February 2, 2015 solved - 6b12 with a disabled disk in the array causes btrfs cache pool to require manual re-balance upon reboot. Remove disabled disk and all is good. Link to comment
garycase Posted February 2, 2015 Share Posted February 2, 2015 Strange behavior => a disabled array disk certainly shouldn't have any impact on the cache !! Link to comment
landS Posted February 2, 2015 Author Share Posted February 2, 2015 agreed - and in all honesty I did do a new config... so it could have been something other than the disabled disk. The only things that changed with new config 1 - disk 3 became disk 2 2 - disk 2 (disabled - no data) removed 3 - disk 4 removed (2 TB WD drive - spot to be taken by a 4 TB hgst) I do not really want to play with bad 3TB disk now - but willing to ship it off if any MOD wants to mess with it. Link to comment
BRiT Posted February 2, 2015 Share Posted February 2, 2015 Strange behavior => a disabled array disk certainly shouldn't have any impact on the cache !! Indeed. Perhaps it was to give the user an option of using part of the cache-drive as a replacement data-drive? Might just have been a data-drive error causes cache-drives bit to not execute... Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.