Strange entry in make_net_recovery logs

[Albert_31]

Hello All,

Can you help me diagonise what the error is all about and how to resolve it..

* lssf: /dev/dsk/c11t: No such file or directory

I did check up the /dev/dsk directory and dont find anything suspicious..there are other device files as well..ex c5, c10 but it always complains of c11t..any hints on what the issue can be.

Regards

Albert

[VK2COT]

Hello,

The physical volume name is incomplete:

/dev/dsk/c11t

One would expect name like:

/dev/dsk/c11t2d3 (or similar)

If we assume that someone did not
create a volume group with incorrect
physical volume, then a faulty physical volume is likely cause of your problems.

Can you run these tests:

lvlnboot -v

vgdisplay -v /dev/vg00

ioscan -fn (and look for NO_HW)

cat /stand/bootconf

If /dev/dsk/c11t... is part of the volume group, then you must attend to it
firstly.

There are some nice references in ITRC
how to attend to a faulty PV in
vg00...

Finally, if you wat to run a comprehensive testing of your server (that includes
many exotic and hard-core tests that I
found extremely useful in my 20 years
of Unix support), here is a Perl script.

Regards,

VK2COT

[rariasn]

Hi Albert,

The message seems to be incomplete. He verifie in the server of ignite the log of the client,

/var/opt/ignite/clients/$HOST_CLIENT/recovery/latest/recovery.log

rgs,

ran

[Albert_31]

Thank you for your replies.. i guess I was not clear enough..my apologies..

a) we run a make_net_recovery on the server and in the recovery.log file, we find the error reported above..There is no message before or after it..i mean a warning which can be linked to the above error..

To troubleshoot..
1)i have run a lssf on all the devices on the /dev/dsk directory and there is no error reported. 2) All the major and minor numbers are in order..
3) There is no device file with "c11t" in the /dev/dsk directory.
4) There are 310 device files in the /dev/dsk directory..c11 is the alternate path to c10 devices..however we dont get any warnings on c10..

Hope it helps in narrowing down the issue..Thank you all for helping me..out of this issue..

regards

albert

[Albert_31]

Hello Everyone,

Guess was not clear enough regarding the issue..

We are getting the above error in the recovery.log file.. to troubleshoot i have done the following and not able to see get anywhere near the issue..

a) lssf on all the device files in the /dev/dsk directory is going fine..no errors at all

b) c10 is teh alternate path of c11..however are getting errors on c11 only.

c) major ,minor, dates , permissions of the device files are in order..

d) There are no other errors in the recovery.log file which we hints of any other issues

e) all VGs are configured properly and all in good order.. have checked (double checked) the same.

any more hints..

regards

Albert

[Steven E. Protter]

Shalom Albert,

Find the device in question.

rmsf it.

insf -C disk

See if it goes away.

This problem needs to be resolved to avoid trouble using the Ignite archive.

ioscan -fnCdisk

Match this against the list Ignite is reporting.

It would appear a disk has been removed or perhaps failed in the system. If it went unnoticed long enough, you might have missed the lbolt in syslog.log

SEP

[Ninad_1]

Albert you have said
"3) There is no device file with "c11t" in the /dev/dsk directory.
"
Also you say "c11 is the alternate path to c10 devices.."

So do you mean to say that c11 are alternate paths to c10 devices but there are no entries for the c11 disk devices in /dev/dsk or /dev/rdsk ?
Then one clarification required - how do you determine that c11 is alternate links to c10 ?
Does your ioscan show the c11 entries and show CLAIMED ?
run insf -e to create the missing device entries and then try again.

Regards,
Ninad

[Albert_31]

Sincerely Apologize for the confusion..

a) The device files for /dev/dsk/c11t* are present in the /dev/dsk directory.

b) lssf /dev/dsk/c11t* works out fine.

c) c10txdy is the alternate path of c11txdy, this I have confirmed with the ioscan, pvdisplay output.


/root# pvdisplay /dev/dsk/c11t9d7
--- Physical volumes ---
PV Name /dev/dsk/c11t9d7
PV Name /dev/dsk/c10t9d7 Alternate Link
VG Name /dev/vg02
PV Status available
Allocatable yes
VGDA 2
Cur LV 73
PE Size (Mbytes) 16
Total PE 539
Free PE 0
Allocated PE 539
Stale PE 0
IO Timeout (Seconds) 180
Autoswitch On

The issue is with make_net_recovery.. I dont know which command actually runs this lssf and which file it is referring to so as to give the error

"lssf: /dev/dsk/c11t: No such file or directory"

If I run the command manually (the same error is reported)

/root# lssf /dev/dsk/c11t
lssf: /dev/dsk/c11t: No such file or directory

The issue is I am not able to determine where is this file or device "/dev/dsk/c11t" is...

[Albert_31]

Hello Everyone..

Tried out the following as well..ie ran the scripts the net_recovery would actually run during the configuratioin collection..

a)save_config --> no error..
b)list_expander --> no error..

does anyone know which command actually runs the lssf command during the net_recovery..??

[Enrico P.]

Hi,
should be make_sys_image

Have you check the lvmtab for error?

ENrico

[Albert_31]

Hello & Shalom To everyone,

Just wanted to update you all that the issue is still pending and the culprit is nowwhere in sight :(..

I have found that it is the save_config (ascii file) command that runs the lssf command to generate the i/o paths in system configuration file, i have played with it today (save_config) file however it did not fail at all (i have captured the values that are input to the lssf command..no wrong devices at all)..it does not give that message at all..

any hints..

regards

albert

[Darrel Louis]

Hi Albert,

Can you please post the following:

cd /dev/dsk
lssf c11*
ioscan -fn
insf -e
lssf c11*

You get the error message when a device file doesn't exists, see below
[gtsx201:root]# lssf c6*
lssf: c6*: No such file or directory

Darrel

[Albert_31]

here the outputs..

[Darrel Louis]

Albert,

What I can see from the output a disk, on c10 and c11, was added yesterday to this server, is that correct?
brw-r----- 1 bin sys 31 0x0be400 Jun 5 12:17 c11t14d4
brw-r----- 1 bin sys 31 0x0ae400 Jun 5 12:17 c10t14d4

Did insf -e create any device files today or is there an error message in your syslog.
Does dmesg report any issue?

Darrel

[Albert_31]

Hello Darrel,

The error is being reported from May 5th onwards.

the new devices were created yesterday since we presented a new lun to the server.

As informed earlier and as in the output, there is no error with lssf command output run manually.

regards

albert

[Darrel Louis]

Albert,

As we can disks has been added since the 4th of May, since that day your ignite is generating the error message.
But did you remove disks aswell?
The disks added on the 4th of may are starting from lun4 till 7.
Did you had any disks from lun0 till 3 on instance 11.
This is purely for my understanding.

Darrel

[Albert_31]

A big THANK YOU to everyone for giving their inputs on this strange issue.

To make a big story short.. THE MESSAGE IS A FALSE MESSAGE.

Have modified the save_config command to print the list of disks being populated to the lssf command output, it comes out clean. Though the message is being logged.

Logged a support case with HP and it comes to light that it is a BUG, solution upgrade to c.6.x..

Thanks once again and happy foruming..

albert

[Albert_31]

Issue with IGNITE UX..