Chris Cain

Were any eSELs generated to the BMC? There should be HTMGT traces in /var/log/syslog or /var/log/messages on the host (collect SOS data). You should also confirm that the opal-prd service...

Looks like OCC generated a log: ``` Jul 10 20:52:03 host opal-prd: HBRT: HTMGT:>>processOccAttn(0x0x7fffb94241d0) ... Jul 10 20:52:03 host opal-prd: HBRT: HTMGT:E>elogProcessActions: OCC1 requested a WOF reset Jul 10 20:52:03...

Was BMC data collected after the failure? (that should have the eSEL data)

opal-prd gathers some error data and sends it to the BMC. The system should be handling the recovery attempts, but if there is a persistent error or need for service,...

What level/commit are you using for hostboot and hostboot-binaries?

FYI, The symptoms in that last opal-prd-fail.log (from yesterday) were the same as the [original failure](https://github.com/open-power/occ/issues/26#issuecomment-510512119). Which is why I had @rbatraAustinIBM confirm the code level.

Let us first confirm that the PM Complex can get reset and still maintain communication with the BMC. Copy the [occtoolp9](https://github.com/open-power/occ/blob/master/src/tools/occtoolp9) script to the host. Boot your system so that...

There were cases where the external HBRT interfaces were called when the OCCs were not started or were reset, so this variable was added to not attempt talking to the...

I am not sure if that last comment had a question or was just a statement. As I said, that HTMGT::processOccStartStatus() function is what sets the iv_occsStarted flag. You do...

Did the OCCs return to active state after the reset? If so, do you have the poll response? If not, are their any PELs / Error logs? What system type...