ok, i had to reboot the unraid with the command "reboot" because the webgui was stuck, and the whole shutdown purpose was to add a new drive.

 

now, it's showing a 'unclean shutdown detected" and if i start the array, it will start the parity check.

 

can i, or should i still add the new disk now?

no one replied, and i had to move on.

 

I added the new disk, and the parity check started, and there was the "format" button, and I did, and after the format was completed, the array was still up, and the parity check continued.

 

however, the parity check is extremely slow (2.5 MB/s), causing to hang while accessing my array via network shares.

 

so I stopped the parity check, i guess I'm not protected now. what's the solution?

If the disk was precleared, you should be still protected.  But that being said, I'm very loathe to do anything like that while a parity check is running.  I would have cancelled it first and then added the drive

 

 

I'm curious as to why the check is so slow.  Can you post a syslog

the disk was precleared before inserting it to my system, i precleared on the side on another system.

 

so now, i stopped the parity-check, how come you are sure that i'm still protected?

 

I can see on the webGUI that there are reads/writes on the parity disk, and it says the parity is valid, even though i interrupted the parity-check.

Because a precleared disk doesn't affect parity at all since its all zeros.  The writes to the parity disk are from the format process of the new disk (updating parity).  Run a check parity to be sure

You probably should have waited a little longer for a response. Tapatalk is showing "3 hours ago" for both of the above messages :-)

 

As far as I know the reboot command reboots the computer but does not actually stop the array before that. You can read the wiki or search the forums on how to cleanly shutdown in command line. I had to do that many times when gui became unresponsive in earlier versions.

 

The parity check was because of the unclean shutdown. As long as there were no write attempts to the array when you rebooted, the parity should be intact. I believe you're still protected, but you should run a parity check to make sure there are no existing errors.

 

With regards to the slow check, make sure there are no loose cables. That's a very common issue. Again I'm not really an expert at this but have been on the same boat before. Perhaps experts can chime in as well.

 

Good luck

I can't run a parity check, it's extremely slow and affecting accessing my array dramatically.

 

attached is my syslog

unraid-syslog-12.04.2015.txt

I can't run a parity check, it's extremely slow and affecting accessing my array dramatically.

 

attached is my syslog

Its slow because you're getting a ton of errors on your drives.  I think the driver keeps resetting the controller card.  If this is a hotswap case, you might want to check that you're not exceeding your powersupply capacity.  If it's not, maybe you're exceeding the capacity, or your slightly loosened some cabling inside somewhere

I can't run a parity check, it's extremely slow and affecting accessing my array dramatically.

 

attached is my syslog

Its slow because you're getting a ton of errors on your drives.  I think the driver keeps resetting the controller card.  If this is a hotswap case, you might want to check that you're not exceeding your powersupply capacity.  If it's not, maybe you're exceeding the capacity, or your slightly loosened some cabling inside somewhere

 

I saw those errors, I had that issue with my old Dell H200 perc card, it had a maximum capacity of 16 drives.

 

Now, my card is LSI MegaRAID 9240-8i.

 

 

btw, those errors are showing only when i'm doing a parity check.

 

my PSU is high, and can handle all the 21 drives I have installed (including the parity disk).

I can't run a parity check, it's extremely slow and affecting accessing my array dramatically.

 

attached is my syslog

Its slow because you're getting a ton of errors on your drives.  I think the driver keeps resetting the controller card.  If this is a hotswap case, you might want to check that you're not exceeding your powersupply capacity.  If it's not, maybe you're exceeding the capacity, or your slightly loosened some cabling inside somewhere

 

I saw those errors, I had that issue with my old Dell H200 perc card, it had a maximum capacity of 16 drives.

 

Now, my card is LSI MegaRAID 9240-8i.

 

 

btw, those errors are showing only when i'm doing a parity check.

 

my PSU is high, and can handle all the 21 drives I have installed (including the parity disk).

I don't have your hardware, but it seems to me that anytime there's task aborts there is something wrong there. 

 

As far as only showing during parity checks thats why I'm inclined to investigate power issues.  That's when you're stressing the system the most.  Could simply be a crappy splitter if you're using them (I avoid them like the plague)

What model PSU?

What model PSU?

 

SeaSonic X750 Gold 750W ATX12V V2.3/EPS 12V V2.91 SLI Ready 80 PLUS GOLD Certified Full Modular Active PFC Power Supply

What model PSU?

 

SeaSonic X750 Gold 750W ATX12V V2.3/EPS 12V V2.91 SLI Ready 80 PLUS GOLD Certified Full Modular Active PFC Power Supply

21 Drives @ maybe up-to 2A maximum = 42A

Cpu would draw at most 8-10A

Mobo's I think are around 3-5A

Fans vary between 1/4 and 1.25 Amp depending on what you have

HBA maybe upto 1 A

 

The P/S can kick out 62A

 

Conservative values on your components is getting close to that value.  If the problem didn't appear until you added the drive, then power capacity or your distribution system is definitely your issue.

 

If you're using splitters to run all those drives, then each of them impose a penalty also.

 

Personally I like to give myself a little bit more headroom.

Also, if all the power is going through one or two separate lines coming from the P/S, then you've also got to remember that each of those lines has a maximum carrying capacity of 18A (with the gauge of wire that's usually used).  So if you're powering all the drives through 2 cables (each that has 4 peripheral molex plugs on them) you're already exceeding the specs of the supply

It is definitely a bad idea to move forward with upgrading an array with an existing problem. You just need to resolve the prior problem, and a new element certainly increases the complexity.

 

My first thought is that you opened the case to insert the new disk and knocked some cabling loose. Try opening and VERY CAREFULLY AND THOROUGHLY check all of the cabling - power and data. I had a similar issue as you that would just not go away. I removed all of the cables and added a drive at a time testing each one. It was a PITA, but I got it up and working. A few days later I started getting resets and red balls. I replaced all the SATA cables with locking cables,. That helped tremendously. Also, I added removable drive cages. Being able to add or swap a disk without risking cable nudge is worth the cost.

 

The power supply angle is a possibility, but with 750 watt single rail I sincerely doubt it. If you were going to have a problem it would be at bootup.