Operating System - OpenVMS
cancel
Showing results for 
Search instead for 
Did you mean: 

DTSS behaving strangely ?

 
Wim Van den Wyngaert
Honored Contributor

DTSS behaving strangely ?

We have VMS 7.3 with decnet 7.3 eco 3 (yes I know, not fully patched).

This Sunday time changed twice while it should happend next Sunday. The strange thing is : all our nodes have the same settings but only a subset of 10% (2 clusters and 1 stand alone node) had the problem.

See enclosure for more data.

Why oh Why ?

Wim
Wim
14 REPLIES 14
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

BTW : we use this version with unchanged settings since 2003 without any problems.

Found another cluster where only 1 node changed time. The node that changed was up 1185 days, the node that didn't change was up 222 days.

Fwiw

Wim
Wim
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

It could be that the same thing happened before but we only enabled time audit in 2007.
Wim
Ian Miller.
Honored Contributor

Re: DTSS behaving strangely ?

could it be that the latest patches update the DST change rules?

Are the all the systems running the same versions ?
____________________
Purely Personal Opinion
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

All running the 2002 version of DTSS. No patches at all since 2003.

Note that time was changed and 3 minutes later again. A bug ? In that case there should be other victems. I checked release notes but found nothing in that direction.

Wim
Wim
Walter Miller_1
Valued Contributor

Re: DTSS behaving strangely ?

The log file clearly shows:

Next TDF Change = 2008-03-30-01:59:59.999+01:00I0.000

So this does not look like a TDF change but a time change brought about by the MASTER servers having an incorrect time. Current configuration on this node only requires two servers to make a time change.

This node is configured as a CLERK - have you checked the MASTER nodes to see if they made a time change?
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

Walter,

The change was at 3:00 sharp ! And the change back 3(!!!) minutes later.

But I checked our 2 providers (again , did it yesterday).

1st : sync at 2:45. Got 3 acceptable date-times. Next sync at 3:45 idem. Nothing happened in between. Nothing in operator log.

2nd : sync at 2:28. Got 3 times. Next sync at 3:28. Nothing in operator log.

None of the 2 nodes connected to the external clocks changed their times. Each node is connected thru a decserver to the 2 external clocks that are about 5 km apart.

There are 2 other nodes not connected to the clock that are server of time. Nothing in their log files and no time change.

Wim
Wim
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

Walter,

On the nodes that did change, operator log says :
DTSS-I-SERTDF DTSS#SERVICE set new timezone differential

So, it did think that it was the change of winter/summer.

Wim
Wim
Walter Miller_1
Valued Contributor

Re: DTSS behaving strangely ?

Yes, that definitely looks like it was a TDF change. Are you running a TCPIP stack and if so - what version? I seem to remember there was a UCX$TDF logical in pre-V5.0 that could possibly conflict with DTSS.
Wim Van den Wyngaert
Honored Contributor

Re: DTSS behaving strangely ?

TCP 5.3 eco 2. On ALL systems, those that have the problem and those that don't.

Wim
Wim