ProCurve Switch 2610-48-PWR (J9089A) & HP Switch 2520-24-PoE(J9138A) Suddenly

[data2]

ProCurve Switch 2610-48-PWR (J9089A) & HP Switch 2520-24-PoE(J9138A) switches keep crashing in the middle of production. We have updated the firmware to the most latest 15.09.0027 (J9138A) & R.11.116 (J9089A) roughly about 2 weeks ago.  They are in a stack, with the J9089A being the commander. I was able to drum up these 2 log entries:

(J9089A)

---- Event Log listing: Events Since Boot ----
I 01/01/90 00:00:04 system: --------------------------------------------------
I 01/01/90 00:00:04 system: System went down: 09/15/16 10:41:24
I 01/01/90 00:00:04 system: TLB Miss: Virtual Addr=0x00000008 IP=0x802a2870 T

(J9138A)

---- Event Log listing: Events Since Boot ----
I 01/01/90 00:01:23 00061 system: -----------------------------------------
I 01/01/90 00:01:23 00062 system: System went down without saving crash
information
M 01/01/90 00:01:23 03001 system: System reboot due to Power Failure

Was hoping someone could help point me in the right direction to get some more information on the crashes or help me interprept the first log. Any help would be appreciated greatly!

[parnassus]

Off top of my head I would try to disable (unconfigure) the "Management Stacking" feature (so then you will be able to manage each switch separately as an autonomous entity with its own Management IP Address) and verify if the issue will still persist with the same actual severity.

Have you any SMNP feature enabled on those Switches?

Also a visual (textual) check of each configuration file could be done to ensure configuration files are in good shape and not (textually) corrupted. Never say never.

[data2]

The biggest thing about this switch is that its still in production, but we do have plans to try some of these ideas over the weekend in a attempt to replicate the problem. Its still currently still on the stack as commander, but they have individual IPs already assigned to them. I have noted that to try this weekend. And we have a SNMP Community enabled for the default "Public". I logged a call with HPE, and both of us took a glance at the textual configuration and found a few things peculiar that didn't look normal, including the flag for "httpd". I remember that I was in the browser at the time of these crashes but can't nail down the exact steps I took in order for it to crash, so one goal we have is to try and figure out the exact steps I took that broke then going to get back to HPE with the steps listing. From my understanding it appears to be a unreported problem with the switch's HTTPs interface. Hopefully I can get some more details as to why it happened to better help the techs. 

Thanks again for your reply!