CPU bound following time change - VMS731_TDF-V0100 didn't work?

[Michelle Popejoy]

Systems: 7.3-1, 8400's, 4100's, DS10, DS20.

Symptom: System(s) become CPU bound following time change by processes such as symbionts, DECevent, and DESTA director.

This issue appears to be resolved by VMS731_TDF-V0100 (or VMS73_TDF-V0100), however...

We installed VMS731_TDF-V0100 in August.

In October following testing which involved changing the date into the future and back again, symptom was experienced (w/ symbionts primarily) on development system.

Just following the time change (EST->EDT)earlier this month, we experienced the problem in production. The symbiont processes were the most noticable, but we were bitten by DECevent and DESTA director as well.

We are working w/ support, but are unable to take development system time right now to duplicate.

The workaround is to stop and restart the various services/applications, but that kind of defeats the purpose of automatic time change. ;-)

Anyone else experience this?

TIA,
Michelle

[Ian Miller.]

I think other patches may help (PTHREAD parahaps?) but I've not seen a definitive list. However I do know plenty of people had problems and made sure hp knew too! I think/hope hp are working on a complete solution

[labadie_1]

I am convinced Ian has spotted it: do you have the last Pthread patch applied ?

[Martin P.J. Zinser]

Hello Michelle,

you are aware about the TDF V2 and V3 patch kits? Most probably hp will want you to install the latest patch kit before going too deep into this.

Greetings, Martin

(P.S. You also will need to install UPD2 with TDF3)

[Ian Miller.]

Although the pthread patch is the key one I think there are others. There are certainly prequisite patches for VMS731_TDF-V0300 including VMS731_UPDATE and VMS731_PCSI

[Michelle Popejoy]

Thanks for the suggestions, guys.

Installed Available
--------- ---------
PTHREAD V2.0 V2.0
TDF V1.0 V3.0 *
UPDATE V2.0 V2.0
PCSI V1.0 V2.0

*We do not have TDF V2 or V3 installed, though we haven't noticed the problems specifically addressed by them (vs. V1).

We won't be able to apply any patches for a while yet, so if anyone had the current patches for these four on 4/3 I'd be curious whether they had the problem or not. Or if you know of another patch that's key.