2 questions about file system

[lazycat]

1. My system alarmed:Logical volume /dev/vg00/SwapVol1 is mirrored but has some stale blocks.Data loss on hardware failure could occur. I ran fsck but no use. I used sam to check swap volume,the system show the information below:
Device File/Mount Directory:/dev/vg00/SwapVol1
Type:Device
Current Priority:0
Mbytes Available:3263 of 3504
Enabled:now/on boot
I don't know what's it mean.
==================
2.I want to extend the file system /dev/vg00/SmsVol using sam. After a long time,it stoped unsuccessfully and return the message:
The command used to extend logical volumes, /sbin/lvextend, has failed.
the command is shown below. The logical volume has not been extended.
vgcfgbackup: Unable to read the physical volume.: I/O error
vgcfgbackup: Invalid LVMREC on Physical Volume /dev/rdsk/c3t6d0
Is that mean a hardware problem, and if it related to the swap volume mentioned in question 1?
==========================
The attachment is the logs of sam. My server is HP J7000 running HP-UX 11.0

[PIYUSH D. PATEL]

Hi,

I may require some more inputs from you to identify the problem.

Can you pls post the output of the following commands:

#vgdisplay -v /dev/vg00
#diskinfo -v /dev/rdsk/c3t6d0
#lvlnboot -v

It is possible that your disk may be having problem and the same disk may be in the root vg-vg00 and the swap filesystem and Smsvol may be on the same disk.

If your swap is mirrored then the mirror harddisk is having problems.

1. Your primary space is enabled and and it gets enabled always when you boot the system.

2. You are trying to extend the smsvol filesystem but since the disk /dev/dsk/c3t6do is having problems, your command is not getting executed.

Pls post the output

Piyush

[lazycat]

Hi,thanks for your reply. I am from China.
Sorry, I cannot post the output because I didn't use these command manually. I just used the sam tool to do the job automatically. And I think the log files may have some useful messages. Have you review it?
Regards
lazycat

[PIYUSH D. PATEL]

Hi,

I have revieved your log files and then only given the answer.

But thro the SAM log it is a bit difficult to find out the exact problem. If you tell the output one by one aslo, then also it can solve the problem.

SAM has only logged in the actions which you tried to take using SAM.
1. Viewing the swap info
2. Extending the logical volume.

The output from the commands which I have told you to execute will be able to diagnose the problem easily.

Piyush

[PIYUSH D. PATEL]

Hi,

If you give a command

#lvsync /dev/vg00/SwapVol1

then it will try to syncronize the mirror copies of the swap lvol. If you get nay error messages during this command then it could be related to the mirror disk.

Piyush

[lazycat]

Hi,Piyush,thanks a lot!

I know what you mean. But I have to wait for a appropriative time to do that because the server is serving for telecommunication and I am not the only administrator.

I will post the output message immediately when I get it. OK?

Regards
lazycat

[lazycat]

BTW: We have another HP server, and it always reboot automatically. I studied the syslog.log, but there have no useful messages.

What should I do to find the root cause?

This server is in a LAN with 2 IP address.

Regards
lazycat

[PIYUSH D. PATEL]

Hi,

Which server ?? Class...Model no ??

It could be due to lot of issues.

tell me the model so that if there are some specific problems with specific server, I can tell you.

Could be due to OS, Appln problem, hardware, lack of swap space, proper patches not loaded etc

Piyush

[lazycat]

Sorry, I forgot it.
The server is 9000/744, HP-UX B.11.00 E
I am a rookie. I just want to know which log files can help me to find the root cause.

Thanks
Regards
lazycat

[Peter Kloetgen]

Hi Lazycat,

please run the following command on your computer:

lvdisplay -v /dev/vg00/SwapVol1

stale blocks mean, there are parts of disk space not available in your swapvolume. Normally this is caused by hardware errors. In my opinion, it would be best for you to replace your disk. It sounds, like you mirrored your system disk, so this would be no problem.

Allways stay on the bright side of life!

Peter

[PIYUSH D. PATEL]

Hi,

Normally for eroor logs you should check the /var/adm/syslog/syslog.log file to check for nay errors or in the same directory syslog.OLD file.

If your server is crashing then you can also check the /var/adm/crash directory. It will contain subdirectories like core.1 and core.2. The files located in these directories cannot be read by vi and these files are normally sent to HP for diagnosis.

Hope this helps,

Piyush

[lazycat]

Thank you all!
The attachment is the output of lvdisplay -v /dev/vg00/SwapVol1.

I wonder if there has a tool to check the harddisk /dev/dsk/c3t6d0?

Regards
Lazycat

[Sridhar Bhaskarla]

Hi,

1. Take make_recovery image ASAP. Keep two tapes.

2. As indicated by Piyush, run lvsync and see if you can get rid of stale extents.

3. Run lvlnboot -v command and note down if there are any errors like "Physical volume does not correspond to this volume group etc".

4. As your lvcreate gave an error already about I/O error, there is a good posibility about the disk is being bad. Run lvdisplay -v on all the lvols in vg00 and see if you find stale entries anywhereelse.

5. To verify if the disk is bad, run the following commands.

diskinfo /dev/rdsk/c3t6d0

Note down the size and bytes per sector. Both should be correct.

dd if=/dev/rdsk/c3t6d0 of=/dev/null bs=1024k count=5

Shouldn't give any error messages.

If the above gives any error, then the disk is bad and you will need to replace the disk. There a lot of thread on how to replace the mirror disk in this forum.

If the disk is responding properly, then you will need to vgcfgrestore the headers. Again, you can search in this forum for restoring a mirror disk.

-Sri

[lazycat]

Ok, I will try!

regards