isr.ior=0*240020.0*e0f82ff80 error

[ahmet_8]

our workstation had an isr.ior error like subject.what does isr.ior mean?and what is reboot after panic?and what does this adress mean?is it about virtual memory?

[Anthony Lennan]

Hi Ahmet,

A "reboot after panic" message will be seen in your shutdownlog file if your workstation has panicked. If a system detects and error (can be a hardware or software error) it can initiate a panic. When this occurs the system will perform a crash dump to disk and reboot.

Determining what caused a system panic can be a very complex task and is generally something that can only be done by a person with an extremely thorough understanding of HPUX's internals.

If you had enough space in /var/adm/crash you should find a folder labelled something like crash.0. Have a look in there and you should see a file called INDEX and hopefully some .gz files.

If you have a support contract with HP and you're lucky enough to have captured a full crash dump you can contact HP and send them the crash dump to have it analysed.

Has anything changed recently in the enviroment (new hardware or software) before the system panicked. If so this would be the first thing I'd look at as being the most likely cause of the panic.

Let us know.

Rgds,
Anthony

[Anthony Lennan]

Hi Ahmet,

Just a couple of other things to look at. Have a look in /var/tombstones and have a look at the file that was created after the server panicked (ie ts99 if it was the last time the system rebooted). Is it different to the files before it (ie ts98 ts97 etc)?

If the file is different it may point to a hardware issue. Once again HP are the best people to analyse these files and they have special tools that can quickly pinpoint the problem if it is indeed a hardware issue.

Also have a /etc/opt/resmon/log/resmon.html and look for any errors that may have occured around the time of the panic.

In regards to the isr.ior error, you will generally see these logged with most kinds of system panics and the number generated doesn't immediately point to the cause (if you're lucky HP may have the exact number recored in their internal knowledge base from previous cases but its unlikely).

Thats all for now. Please have a look at those files I've mentioned and let us know.

Cheers,
Anthony

[melvyn burnard]

These are pointers to addresses in the kernel where a possible error occured.
Your system appears to have had a panic, you need to:
1) look at /etc/shutdownlog to see what it says there.

2) See if you have a /var/tombstones directory, and look in the ts99 file and check if there are any timestamps matching the time of the panic as this may tell you you have a hardware error.

3) check to see if /var/adm/crash has a crash directory created around the time of the panic, and if so what is in the INDEX file following the entry panic string:
and then log a call with your local HP Response Centre and have the dump analysed.

[Kent Ostby]

The isr.ior "panic" will fall into one of three categories:

1) TOC -- This happens if someone generates a TC/TOC dump via GSP, CM, or the core-io board. If the box just rebooted itself then this isn't your cause.

2) Hardware error -- Look INSIDE the /var/tombstones/ts99 file and for each CPU you should either see a valid timestamp or no valid timestamp. If you have at least one valid timestamp inside then you had a hardware error.

3) ServiceGuard TOC -- depending on your OS revision, a TOC caused when two or more MC/ServiceGuard nodes cannot communicate, may show up as an isr.ior panic. If you don't run ServiceGuard then this obviously doesn't apply.

Best regards,

Kent M. Ostby