Had Unraid for about 2 months now.

 

Did preclear on 2X2TB drive (2 cycles each) and no problems. One is parity, one is data disk 1. Did parity check AFTER I setup as suggested, 0 errors. Re purposed a 1TB drive and did 2 preclear cycles without any problem and added the 1TB drive as disk 2.

 

Decided to run a parity check after just adding to array like most people suggest. Let it run overnight and it had 7300+ parity errors.

 

Ran parity check again and this morning 0 errors? I have not changed or added ANYTHING between the 2 parity runs. What gives? What does it mean or what should I do?

 

Also I recall reading somewhere (could find it again to save my life) about different types of parity checks. One is something like Read and NOT fix, the other is Read and fix. I don't know which one I have but I haven't done ANYTHING special to setup. So whatever the "Default" way is...I have that.

 

Should I be worried on this one guys?

Had Unraid for about 2 months now.

 

Did preclear on 2X2TB drive (2 cycles each) and no problems. One is parity, one is data disk 1. Did parity check AFTER I setup as suggested, 0 errors. Re purposed a 1TB drive and did 2 preclear cycles without any problem and added the 1TB drive as disk 2.

 

Decided to run a parity check after just adding to array like most people suggest. Let it run overnight and it had 7300+ parity errors.

 

Ran parity check again and this morning 0 errors? I have not changed or added ANYTHING between the 2 parity runs. What gives? What does it mean or what should I do?

 

Also I recall reading somewhere (could find it again to save my life) about different types of parity checks. One is something like Read and NOT fix, the other is Read and fix. I don't know which one I have but I haven't done ANYTHING special to setup. So whatever the "Default" way is...I have that.

 

Should I be worried on this one guys?

 

YES!!!  But you also failed to provide the members with very much information.  You need to state which version of unRAID you are using and to attach a syslog for people to be able to diagnose  your problem.  (http://lime-technology.com/forum/index.php?topic=4640.0)

Had Unraid for about 2 months now.

 

Did preclear on 2X2TB drive (2 cycles each) and no problems. One is parity, one is data disk 1. Did parity check AFTER I setup as suggested, 0 errors. Re purposed a 1TB drive and did 2 preclear cycles without any problem and added the 1TB drive as disk 2.

 

Decided to run a parity check after just adding to array like most people suggest. Let it run overnight and it had 7300+ parity errors.

 

Ran parity check again and this morning 0 errors? I have not changed or added ANYTHING between the 2 parity runs. What gives? What does it mean or what should I do?

 

Also I recall reading somewhere (could find it again to save my life) about different types of parity checks. One is something like Read and NOT fix, the other is Read and fix. I don't know which one I have but I haven't done ANYTHING special to setup. So whatever the "Default" way is...I have that.

 

Should I be worried on this one guys?

 

Most probably (but depend on more data you need to provide):

 

Parity errors mean that the data disk that is set up to make sure you can loose a disk has contents that do not correspond with what is actually on the disk...

 

Check SMART results on your drives, if they are ok and you also see no red balls on your drives in unraid then it could be  that you have reset your system a couple of times without a clean shutdown, this could cause parity errors (AND data corruption by the way).

 

The moment you did the parity check you probably had the "correct parity" check on, this made sure that the parity was rewritten to "fit" to what is actually on the drive (in other words: f you had dat errors on your drive now the the parity drive is set up to recreate those same parity data errors the moment you exchange a drive..

 

You need to find out if this was a one time event or if there is something more structuraly going on, if you redo a parity check and nothing is found, and if you try it again, and zero errors again next week, then *something* happened that gave you the errors but probably it is not a failing drive (check smart results) or a hardware incompatibility.

 

If you rerun parity check and it finds errors AGAIN, then do not trust your array with any more data,  copy your syslog to a safe location and ask for expert help on the forum..

Ok....Followed this post (http://lime-technology.com/forum/index.php?topic=4640.0)to get you guys a "syslog file"

 

Seems like a foreign language to me and I appreciate all the help to understand what is going on.

 

And as Helmonder pointed out I may have not had "clean shutdowns" without knowing it. I recall reading somewhere that you can make something to do a clean shutdown when pushing the power button? Is there more than one way to do this or just a catch all program that I just need to make work? And how to do it?

 

Thank you guys for spoon feeding a noob like me

 

syslog.txt

It sounds like some errors were detected and fixed in your first parity check. I would run smartctl on the drive with the errors (assuming there's only one).  Please post the results of the SMART test.

 

And to answer your original question: parity errors are discrepancies between what the system expects to find and what it does find. When it runs a parity check, each bit is a 1 or a 0. The corresponding bit on the parity drive is derived from whether the "sum" of all the bits are even (0) or odd (1). That's why you can rebuild a lost drive. The system knows what every bit of on a lost drive has to be based on the bits on the other drives. That's also why you can only recover from one bad drive at a time. Two bad drives means that either could hold either value.

 

And as Helmonder pointed out I may have not had "clean shutdowns" without knowing it. I recall reading somewhere that you can make something to do a clean shutdown when pushing the power button? Is there more than one way to do this or just a catch all program that I just need to make work? And how to do it?

 

 

OK, it not really explained any place that I have found but the proper way to shutdown the server is from 'Main' page on the user interface.  First, stop the array.  Once the array is stopped, you should be presented with a new set of options-- one of which is to shut the server down. 

There is a powerdown script, which will perform the necessary steps for a clean shutdown. Generally to be used as a last resort. I believe it's part of the Unmenu packages.

Ok....Followed this post (http://lime-technology.com/forum/index.php?topic=4640.0)to get you guys a "syslog file"

 

Seems like a foreign language to me and I appreciate all the help to understand what is going on.

 

And as Helmonder pointed out I may have not had "clean shutdowns" without knowing it. I recall reading somewhere that you can make something to do a clean shutdown when pushing the power button? Is there more than one way to do this or just a catch all program that I just need to make work? And how to do it?

 

Thank you guys for spoon feeding a noob like me

 

Make it simple for yourself:

 

There is only one way to shut down the system: go to the main page, stop the array, and then shutdown the array.. Only then are you sure nothing is forgotten.. If you have shut down by power button that could well be the reason for your parity errors...

 

Do two things:

 

- Check SMART reports: if one of your drives are showing failure signs (ask in forum if in doubt) then replace drive;

- If SMART ok (or after drive replace) run another parity check, and then another one in a week. If the second parity check does not give any errors you are fine.