This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
Disk test - Does it work or is one of my disks faulty?
Firmware 6.9.4 Hotfix 1
2 x WD40EFRX
Getting 20MB/s from a volume that has been balanced and defragmented looks like it might be something a WD drive does when it starts to fail.
Have ruled out it being AV (it's off) a memory leak etc. (booted up fresh), snapshots (the disks are otherwise idle) and any networking issue (I tried plugging in a USB stick and copying to that via the admin page and copying was just as slow.
So I started a disk test.. it says a disk test is running but the activity light doesn't appear to show any activity and the drives are not making any other sounds:
[18/11/14 19:20:19 WET] notice:volume:LOGMSG_DISKTEST_STARTED_VOLUME Disk test started for volume data. .. [18/11/15 00:49:41 WET] notice:system:LOGMSG_SYSTEM_HALT The system is shutting down.
How long should a disk check take, should there be any indication of progress and is it broken on the latest ReadyNAS OS? Any other tips for slow read speeds (even when a FAT32 USB3 stick is plugged in the back locally)?
Does things like bitrot detection / checksum / quota make a difference to speed?
Re: Disk test - Does it work or is one of my disks faulty?
Thanks.
I've thrown some more time at this and have determined both disks are failing but because they don't get much usage before the ReadyNAS is shut down for each evening the issues were masked.
I have performed a DIsk Test on 2 other ReadyNASes I own and they have all started correctly scanning. The 312 where the Disk Test fails shortly after issuing it now also has 1 drive that is showing a "Current Pending Sector Count" of 2.
Interesting that the drives appear healthy until a full disk test and that the WD firmware has a go-slow with no other errors triggered if the disks are starting to misbehave.. I suspect the raw throughput test worked as I only ran it on a bit of the drive.. and the Scrub worked (albeit slowly) because the WD isn't overly noisy with SMART logging.. It's only the Disk Test that shows up serious errors. I will make sure to keep my ReadyNASes more routinely disk tested now!
The disks are about 2 months out of the 3 year warranty (always the way isn't it)
I'll come back to update if the raw throughput recovers once my new drives arrive 🙂
Re: Disk test - Does it work or is one of my disks faulty?
My RN526 runs 4x6TB WD60EFRX in RAID-5 (XRAID). The last disk test took about 16 hours. That was with 6.9.3 though. NasTester shows about 290 MB/sec write speed and 360 MB/sec read speed (over 10GbaseT).
If you enable ssh, you could look at memory loading and also separate the filesystem performance from the network.
Re: Disk test - Does it work or is one of my disks faulty?
Thanks.
I've thrown some more time at this and have determined both disks are failing but because they don't get much usage before the ReadyNAS is shut down for each evening the issues were masked.
I have performed a DIsk Test on 2 other ReadyNASes I own and they have all started correctly scanning. The 312 where the Disk Test fails shortly after issuing it now also has 1 drive that is showing a "Current Pending Sector Count" of 2.
Interesting that the drives appear healthy until a full disk test and that the WD firmware has a go-slow with no other errors triggered if the disks are starting to misbehave.. I suspect the raw throughput test worked as I only ran it on a bit of the drive.. and the Scrub worked (albeit slowly) because the WD isn't overly noisy with SMART logging.. It's only the Disk Test that shows up serious errors. I will make sure to keep my ReadyNASes more routinely disk tested now!
The disks are about 2 months out of the 3 year warranty (always the way isn't it)
I'll come back to update if the raw throughput recovers once my new drives arrive 🙂
Interesting that the drives appear healthy until a full disk test
Generally speaking, the disks won't detect problems unless something fails - usually a read or a write request to specific sectors. So that really isn't that surprising.
In addition to the disk test, a scrub also exercises all the disks in the volume pretty thoroughly.
Interesting that the drives appear healthy until a full disk test
Generally speaking, the disks won't detect problems unless something fails - usually a read or a write request to specific sectors. So that really isn't that surprising.
In addition to the disk test, a scrub also exercises all the disks in the volume pretty thoroughly.
True, I think my point was that ReadyNAS sees "Reallocated Sector" info in the SMART but didn't directly warn until I started digging into the full support logs.
