Hi All,

 

I have a problem with a fresh install of unRAID.

I have 6x2TB drives to give 10TB with parity and a separate 320GB cache which isn't configured for use as a cahce drive, just to hold plugin data.

 

When i open up the webGUI, all disks shows a very high read count, as seen below:

 

I'm currently writing to disk5 as a timemachine backup, and transmission is working on a share that is set to use disk1.

 

Can anyone point me in the right direction to identify why this is happening?

All plugins, except transmission are disabled, as i'm sure it isnt transmission that is causing it as i have tried disabling it and the reads keep increasing.

 

Thanks,

Austin

Just tried a reboot, and cancelled the parity check upon reboot.

Even after cancelling the parity check it's still giving lots of reads...

 

I'm going to run the parity check with no other activity running as I imagine this is due to the parity check not having been done and having to read all disks to calculate parity as I write data to the array.

Just tried a reboot, and cancelled the parity check upon reboot.

There should not be an automatic parity check unless you did not shut the array down properly. When you stop the array, you should get the option to shutdown or reboot. That is the only way you should be rebooting the array until you read up on some of the other available options. If the array is taken down without being stopped, there is a high likelyhood that the parity disk is wrong, and will not be able to recover from a failed disk. I would do a full correcting parity check ASAP.

.

Since you have a new setup: did you turn OFF "spin up groups" ?  If it is on then more drives will spin up then the ones you are actively using..

Spin-up-groups will not result in "reads" of the disks.  (also, unless the disks share a spin-up-group name, they do not spin up.)

 

Reads will certainly occur during parity calculation and parity checks, but if you cancelled the check, then no reads should be occurring UNLESS you have an application of process on your LAN (or as an add-on) that is accessing the drives.

 

Easiest way to identify what is accessing the disks is to use

inotifywait

as described in this post. http://lime-technology.com/forum/index.php?topic=25973.msg226486#msg226486

 

Another method is t turn on disk buffer debugging, as described in the second link in this post:

http://lime-technology.com/forum/index.php?topic=27162.msg238578#msg238578

Just dn't frget to turn this off, since it will eventually fill the syslog and likely crash your server.  (don't do it when the parity check is occurring, otherwise you'll get millions of messages as it checks/calculates parity.)

 

Joe L.