iLO3 Unresponsive on Proliant BL460C

[William Graham]

Hi I recently posted on another thread requesting Beta Firmware that fixes the problem where iLO3 will respond to pings, but not give access to iLO Web Page or SSL.

Our company uses Foundstone, we have turned it off and still the same issues.

The iLO firmware is 1.16, the most recent version on HP Support & Drivers.

On another thread, I found 1.20 Firmware, I applied this to check if it fixed the issue I was having. After applying the firmware and power cycling the server, it came back up, iLO did not respond to pings, and did not give access to iLO Web Page or SSL.

Any ideas?

[Oscar A. Perez]

Tell me if I understand correctly, you flashed iLO3 with the new version 1.20 and iLO3 was up for a while then, suddenly became unresponsive, right?

How many servers with iLO3 do you have? Are they all showing the same symptom? Is the OA marking these blades with a RedX?

BTW, what OA version do you have? Any Virtual Connect modules?

[William Graham]

Oscar, Thankyou for you reply.

I flashed iLO3 with 1.20, and rebooted the server. Immediately it became unresponsive.

We have 2 x C7000 Chassis, each chassis has 14 bl460c G2 Blades, they run iLO2 and work perfectly. We recently purchased 4 x bl460c G7 blades, they use iLO3 and are all having the same issue.

When we log in to the Onboard Administrator on each Chassis, no errors, or red X's are present, in fact, we are able to use the Virtual Power features to reset, cold boot the blade (not sure whether this uses iLO or not).

OA is at version 3.21.

We have 4 x HP 1/10Gb VC-Enet Modules per chassis, so 8 in total. They are version 3.15 firmware.

[Oscar A. Perez]

William,

If you re-insert the blade with that iLO3 that you flashed to 1.20, would iLO3 come back? If not, I believe the firmware flash process failed and that iLO3 should be in recovery mode.

In any case, please call HP support and log a case. Then post the case number here so, I request our team to pull it.

[William Graham]

Oscar,

I powered down the server and pulled the blade from Bay 7. I left it out for about a minute then pushed the blade back in again. iLO 3 is still unresponsive. This is only when FW 1.20 is applied, when I put 1.16 back on again, it pings, but nothing else.

At the minute, I have left FW 1.20 on.

The HP Case # is 4626296914-466, the call has been open for a while. It has been logged by my colleague, so the name on the call may not be William Graham.

Thanks for your help.

William

[Erik Språng]

I experienced the same behaviour but on SL390s bricks.

It turned out to be the default gateway parameter which was not configured correctly. Seems like ilo will silently ignore packets sent from other subnets than its own or the default gateway. Firmware 1.16 did reply to pings from other subnets while 1.20 does not.

Hope this helps!

/Erik

[William Graham]

Erik,

Thanks for the reply. It's funny that you should mention that.

The IP Address, Subnet Mask and Default Gateway is meant to be set using EBIPA, in the Onboard Administrator, but when I check the individual iLO3 settings on the server, the default gateway is blank.

How were you setting the default gateway? Were you manually setting this?

Thanks

William

[Erik Språng]

William,

I set it using the RBSU (pressing F8 at boot). I also tried setting it using the HP Smart Start disc, but it would not update the settings correctly.

Smart Start would however view lots of other information the RBSU would not display. For instance the MTU which was set to 5220. That had me on a wild goose chase for awhile. :)

I believe the setup is different in the SL series "bricks" compared to the BL series blades as the SLs are individual machines only sharing power and cooling. There's no common management for the chassis as it is for the blade enclosures. So your experience may vary.

/Erik

[William Graham]

Erik,

You're a star!!! You certainly put me on the right track.

When I tried to set Default Gateway on EBIPA, it was blank on the servers ILO Configuration.

When I manually tried to set the same information in ILO , it would not take the gateway.

Upon further investigation, I noticed that the ILO IP and Gateway were not on the same network. I spoke to our network team who advised they had recently changed something and had given me those addresses in error.

A simple change of Subnet Mask, and the iLO addresses started responding to pings and iLO now works 100%.

Many thanks for your input, it put me in the right direction.

[Erik Språng]

Glad I could help.

Hope you had a nice Easter!

/Erik