Highlighted
hne-dc
1 Nickel

R820 How to recover from a bad iDRAC firmware update?

Jump to solution

I have a pair of identical R820 systems.  I upgraded the iDRAC on the first unit from 2.61.60.60 to 2.63.60.62 without any problem.  I then attempted the same from the 2nd unit, but the iDRAC would no longer respond afterwards.

The update dialog on the 2nd unit went like this and seemed to finish ok.

....

iDRAC
The version of this Update Package is newer than the currently installed version.
Software application name: iDRAC
Package version: 2.63.60.62
Installed version: 2.61.60.60
....
Device: iDRAC
Application: iDRAC
Update Successful.
The update completed successfully.

...

I then did a cold boot of the system, the Lifecycle Controller shows as DISABLED, and later the message "iDRAC initialization error" appears
Management Engine Mode: Active
Management Engine Firmware Version: 0002.0001
Patch: 0005
Build: 00A7
F1 to continue, F2 to run setup.  

The OS will boot, fans on full blast.  I tried running the firmware update a 2nd time, but it just times out.

I've tried recovering from a DRM created SMART ISO and a PER820 Platform Specific Bootable ISO, but both of those hang during inventory collection.

I was able to boot into a SLI ISO.  The BIOS and H320 controller firmwares are already the latest.  However, DSU refused to update the iDRAC complaining of a "partial inventory".

I've already tried powering the system off and "flea" draining the power from the iDRAC.

What's the next thing to try to recover the iDRAC?  SUU?  Any other suggestion or more detailed steps to try?

0 Kudos
1 Solution

Accepted Solutions
hne-dc
1 Nickel

Re: R820 How to recover from a bad iDRAC firmware update?

Jump to solution

And... I fixed it!

"PowerEdge: iDRAC Recovery Procedure with firmimg.d7" was key to re-enabling Lifecycle services.  The imediate noticable improvement was the iDRAC would report Done right away after the memory was configured.  I would still get the iDRAC initialization failed message later on.  Hit F10 to enter the Lifecycle services and  reset the iDRAC configuration to default.  The rest is a few reboots and re-configuring the iDRAC (or System) settings.

To get the firmimg.d7 file, download the iDRAC firmware for Windows.  It can be extracted using "unzip -L <file.exe>".  Unzip it in a subdirectory because it does not create its own.  The firmimg.d7 file is under the payload directory.

I did clear the NVRAM at one point with the jumper, but that didn't seem to help.  I'm not sure if it contributed to the final solution.

0 Kudos
3 Replies
hne-dc
1 Nickel

Re: R820 How to recover from a bad iDRAC firmware update?

Jump to solution

Replying to myself...

"PowerEdge: iDRAC Recovery Procedure with firmimg.d7" using an SD card would seem to be the best bet.  However, how can you extract the firmimg.d7 from the Linux DUP?

 

ref: https://www.dell.com/support/article/us/en/04/how12633/poweredge-idrac-recovery-procedure-with-firmi...

0 Kudos
Charli1232
1 Copper

Re: R820 How to recover from a bad iDRAC firmware update?

Jump to solution

I always get Dell servers with iDRAC with dedicated NIC so that the iDRAC itself uses a static IP address.

Then the iDRAC can be used for many purposes....

- view console of ESXi

- email notification of failures and temperature 

- remote power up of the physical server 

-<Non-public info > 

0 Kudos
hne-dc
1 Nickel

Re: R820 How to recover from a bad iDRAC firmware update?

Jump to solution

And... I fixed it!

"PowerEdge: iDRAC Recovery Procedure with firmimg.d7" was key to re-enabling Lifecycle services.  The imediate noticable improvement was the iDRAC would report Done right away after the memory was configured.  I would still get the iDRAC initialization failed message later on.  Hit F10 to enter the Lifecycle services and  reset the iDRAC configuration to default.  The rest is a few reboots and re-configuring the iDRAC (or System) settings.

To get the firmimg.d7 file, download the iDRAC firmware for Windows.  It can be extracted using "unzip -L <file.exe>".  Unzip it in a subdirectory because it does not create its own.  The firmimg.d7 file is under the payload directory.

I did clear the NVRAM at one point with the jumper, but that didn't seem to help.  I'm not sure if it contributed to the final solution.

0 Kudos