Hi guys,

 

2 Days ago I saw that my server's link speed dropped from 1Gbps to 100Mbps, this has happened many times before and a simple reboot usually resolves it and the link speed reverts back to 1Gbps. This time I rebooted the server, I could see the shares on my mac, but couldn't access the web GUI. So then I hard shut down the server thinking it might be stuck on start up or something, then boot back up, I can telnet to it but can't see the GUI anymore. I have modified the hosts file to point to the IP, and it doesn't matter if I use the IP or the hostname. I can see it connected through my router though so I know it's in the network.

 

It's been fine for over a year now, I do regular parity checks and never came up with an error (knock on wood).

 

Any way, my release is unRAID Server Version 5.0-beta11, and here is the syslog. I need some ideas on what's going on.

 

Thanks a lot,

 

Abdulla

syslog-2012-11-18-2340.txt

It looks like there may be 2 problems. A drive is giving an error and the NIC is unstable. Try a new Ethernet cable. Use CAT 5e or CAT 6. (CAT 5 cable will only go 100Mb.)

Hi guys,

 

2 Days ago I saw that my server's link speed dropped from 1Gbps to 100Mbps, this has happened many times before and a simple reboot usually resolves it and the link speed reverts back to 1Gbps. This time I rebooted the server, I could see the shares on my mac, but couldn't access the web GUI. So then I hard shut down the server thinking it might be stuck on start up or something, then boot back up, I can telnet to it but can't see the GUI anymore. I have modified the hosts file to point to the IP, and it doesn't matter if I use the IP or the hostname. I can see it connected through my router though so I know it's in the network.

 

It's been fine for over a year now, I do regular parity checks and never came up with an error (knock on wood).

 

Any way, my release is unRAID Server Version 5.0-beta11, and here is the syslog. I need some ideas on what's going on.

 

Thanks a lot,

 

Abdulla

 

What IP address are you pointing your browser at?  It appears to me that your server address is 192.168.1.5  (http://192.168.1.5) 

 

You also need to find how to do a proper shutdown on your system.  Just turn the power off generates more problems than any other cause.  Read here for more information:

 

http://lime-technology.com/wiki/index.php/UnRAID_Add_Ons#Power_Down

This is often cause by a plugin failing to start up properly. Try moving the plugins out of /boot/config/plugins over you network shares or telnet and reboot. If that fixes it, install them one at a time and see which one is blocking the process and troubleshoot that on the appropriate plugin thread.

 

whiteatom

If it was a problem with the cable, how am I able to access it via Telnet ? I'll get another cable and try tomorrow.

 

I don't have a plugins directory, I have unmenu installed I think but it's on the main directory, should I move that somewhere else or delete it ?

 

How can I fix this issue and get my server back online ? So I can see which drive has the problem and I can replace it.

 

Thanks guys,

EDIT : OK, It turned out to be an issue with the cable. Thanks dgaschk, whiteatom and Frank1940.

 

P.S. You said that there's a faulty drive. I did a parity check and everything looks fine. Do you know which one is it ? Should I swap it with another one ?

 

Marking this as resolved.

 

Abdulla

Check SMART reports. At least one drive has pending sectors and need to be rebuilt.

I did check the SMART reports, and every single drive has an overall status of (PASSED). Can you be more specific on what I should check ?

 

I'm really sorry about this, I'm not an expert in unRAID.

The "Current_Pending_Sector " RAW value indicates the number of unreadable sectors. All sectors on all disks must be readable in order to rebuild a failed disk. Pending sectors will be resolved on the next write to the sector. Rebuilding the disk should reduce the number of pending sectors to zero. If there is more than one disk with pending sectors then ask for further advise.

Oh I didn't know how to read the S.M.A.R.T. reports, I see one drive with 6 unreadable sectors.

 

When you say rebuilding the disk, do you mean replace it with another drive ? Wouldn't the disk just reallocate those unreadable sectors with some of it's spare sectors ?

 

Let me know what I should do. Thanks tremendously for your help.

 

 

If you have a spare pre-cleared disk, use it now and then pre-clear the disk with pending sectors. If the pending count goes to zero the drive becomes your new spare. If there is no spare then follow this procedure:

 

Go to the Main page

Stop the array.

Un-assign the disk.

Start the array.

Stop the array again.

Re-assign the disk.

System should indicate there is a "new" drive. Check the confirmation box and click Start to start a parity-reconstruct of the disk.

 

The pending count should be zero. If it's not, then the disk needs to be un-assigned and pre-cleared (possibly repeatedly) until the pending count is zero. Then reassign and rebuild.

 

Just to confirm, I wouldn't lose any data doing this, right ? 

No. This is like replacing or upgrading a disk.

before you do anything, do a parity check.  This will ensure the other disks you are relying to re-construct the one with the pending re-allocation sector are all working.

 

Otherwise, you might end up with unknowingly attempting to reconstruct from a disk that is failing itself.

 

 

Man, I just did a parity check on the 19th and it was error free. I still haven't done what dgaschk suggested. Just did another check yesterday like you suggested and it came up with 300~ errors

 

I'm doing another parity check again as we speak. What should I do now that I know there are errors ? I'll get the spare precleared and ready, I should've done it a while ago but I'm just so lazy

 

Thanks for the tip Joe, waiting for your answer on this.

The errors are most likely caused by the disk with unreadable sectors.  Do not run a correcting check! The parity disk is correct. The disk with pending sectors is the issue. Follow the procedure below.

I didn't know and did a correcting parity check It was checked and I didn't uncheck it before doing the parity check. Now it returns 0 errors.

 

Should I still do the procedure you told me about on the first page ?

I didn't know and did a correcting parity check It was checked and I didn't uncheck it before doing the parity check. Now it returns 0 errors.

 

Should I still do the procedure you told me about on the first page ?

 

The disk with unreadable sectors still needs to be rebuild. But the unreadable sectors may not receive correct data.