Monitoring of Subsystems

[Donovan Israel Murray]

Hello Guys. I need some help.

I have a RP5470 running hpux 11.0 with redudant power supply, dual 10/100 NIC in addition to the NIC on the core I/O. This system is connected to a SC10 via a 4si raid controller.

The SC10 has 8 18 GB disks installed and redundant power supplies and fans. I also have a DLT 1/9 autoloader installed.

I have been asked to demonstrate the monitoring aand redundancy features of the installed system. In particular my superiors want to see what happens when a power supply is removed from the server while it is up and running, what happens when a hard disk is removed from one of the configured RAID 5 logical disk, things like that.

Well, with the system up and running I removed a power supply from the server and nothing noticable happened. However when remove a power cord from the back we got a notification from the GSP card.

When I started irm to look at the disks and logical drive under the control of the 4si card it indicated that all the drives were online. I then removed one of the drive with the system up and I received no notification of a failure. irm was also reporting that the drive was still online.

Well of course my superiors were not impressed. Can someone help me to answer some of the natural questions thatcomes out of my experience?

IS the 4si capable of RAID expansion and or RAID migration?

Thank you all in advance

Rgds

[Michael Tully]

I can't help with your disk subsystem as I am not familiar with that unit. As far as the removing of the power cable, creating an entry in the GSP logs is what would be expected. If it didn't create the entry, you wouldn't know there was a failure right....

[Sridhar Bhaskarla]

Hi,

I strongly suggest you configure EMS subsystem on your server. It can send you alarms on most of the components.

/etc/opt/resmon/lbin/monconfig will allow you to configure events. If EMS is already enabled by chance, look at root's mails. You should have some messages out there. Otherwsie, you can configure EMS to send snmp traps or mail.

Explore EMS.

-Sri

[Steven E. Protter]

You should see output in dmesg
and /var/adm/syslog/syslog.log indicating an lbolt and related to what you did.

The fact that the system continues running shows that you spec'd and configured it properly.

Not much of a clue on the RAID part. Sorry.

P