Proliant DL380e Gen8 keeps crashing (PSOD) after upgrade to HP customized VMWare ESXi 5.5 U2

[ErikV1991]

Bad news, this moring the server crashed with the same PSOD.

add VMware boot flag timerEnableTSC = false is not the soloution.

[cykVM]

Thanks for the feedback, ErikV.

Just to be sure: After setting the timerEnableTSC = false you rebooted the server for the boot flag to go into effect?

[ErikV1991]

Hi,

No i did not reboot, but i checked en the setting is still off.

So now it's rebooted i think afhter the PSOD, but i reboot it now.

Let you know if it stay's up untill tomorrow.

[cykVM]

Changed boot flag go into effect only if VMware is rebooted, it's kernel flags which can't be changed live.

So we will see what happens now ...

[ErikV1991]

Hi,

This moring the same PSOD, after change the setting i rebooted the server but no effect.

Any ideas anyone ?

[cykVM]

Thanks for the feedback. I presume the only available workaround (and this is only a workaround!) is to go back to previous version with the SHIFT+r method.

Instead of doing this you may also use the HP customized VMWare 5.5 Update 1 version, which runs fine for me on the same hardware.  See https://my.vmware.com/de/web/vmware/details?downloadGroup=HP-ESXI-5.5.0U1-GA&productId=353 for that.

There is no other option, I guess, until HP and/or VMware fixes this.

[JaaM]

For now, there must be some kernel/drivers issue...

If i want to upgrade, i'll get ESXi 5.5.0 U1...

BUT, i'll let you know if i see something with U2 and all recommendations given before...

Thanks

[cykVM]

Just discovered this: http://h20566.www2.hp.com/portal/site/hpsc/template.PAGE/public/psi/swdDetails/?sp4ts.oid=5269386&spf_p.tpst=swdMain&spf_p.prp_swdMain=wsrp-navigationalState%3Didx%253D0%257CswItem%253DMTX_561321363f664c7a8f75354c13%257CswEnvOID%253D4166%257CitemLocale%253D%257CswLang%253D%257Cmode%253D3%257Caction%253DdriverDocument&javax.portlet.begCacheTok=com.vignette.cachetoken&javax.portlet.endCacheTok=com.v... (new VMWare 5.5 driver for B120i / B320i Smart Array controller - hpvsa version 5.5.0-90 [2014.09.11])

and this: http://h20566.www2.hp.com/portal/site/hpsc/template.PAGE/public/psi/swdDetails/?sp4ts.oid=5269386&spf_p.tpst=swdMain&spf_p.prp_swdMain=wsrp-navigationalState%3Didx%253D0%257CswItem%253DMTX_8372c55483b9432abd53d91951%257CswEnvOID%253D4166%257CitemLocale%253D%257CswLang%253D%257Cmode%253D3%257Caction%253DdriverDocument&javax.portlet.begCacheTok=com.vignette.cachetoken&javax.portlet.endCacheTok=com.v... (new Firmware 2.02 for iLo4)

both released on 06 Oct 2014. Maybe this helps?

P.S. for 5.1.0 Version of VMWare the new hpvsa driver can be found here: http://h20566.www2.hp.com/portal/site/hpsc/template.PAGE/public/psi/swdDetails/?sp4ts.oid=5269386&spf_p.tpst=swdMain&spf_p.prp_swdMain=wsrp-navigationalState%3Didx%253D%257CswItem%253DMTX_e036863187434e4c9d32533b87%257CswEnvOID%253D4141%257CitemLocale%253D%257CswLang%253D%257Cmode%253D%257Caction%253DdriverDocument&javax.portlet.begCacheTok=com.vignette.cachetoken&javax.portlet.endCacheTok=com.vig...

Both updates are RECOMMENDED and do not directly address the issues with VMWare 5.5 U2.

[cykVM]

The new hpvsa version 5.5.0-90 driver seems to be kind of buggy. I installed it yesterday morning, install went fine and 5.5.0-86 was replaced followed by a host reboot, but after that the throughput on accessing Windows shares between the VMs went down massively. I recognized it on yesterday evening/night backup as the data backup which takes about 2-3 hours usually now was still running this morning and would take another 8-10 hours approximately. Throughput on backup to LTO4 tape drive is usually at about 4 GB/min, with the 5.5.0-90 hpvsa driver it went down to about 0.3 GB/min - that's about 10% of the former speed.

I just went back to 5.5.0-86 driver and throughput is back to normal 4 GB/min.

So 5.5.0-90 hpvsa is not recommended on this configuration. I'm using Backup EXEC 2010 R3 on Windows 2008 R2.

What's going on @HP? ;-) Can't remember joining the customer-beta-test-program ...

[cykVM]

And an additional 2 users with these issues posted in VMWare forums: https://communities.vmware.com/thread/491027?start=15&tstart=0