| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 648.1 | IPMT, Please | XDELTA::HOFFMAN | Steve, OpenVMS Engineering | Tue May 27 1997 11:33 | 4 |
|
Elevate this via IPMT, please -- I'd not want the Paris subway running
an ECO OpenVMS didn't directly provide nor support...
|
| 648.2 | may not be the problem | EVMS::MORONEY | vi vi vi - Editor of the Beast | Tue May 27 1997 18:46 | 3 |
| Also check the setting of the SHADOW_MBR_TMO SYSGEN parameter.
-Mike
|
| 648.3 | This is a "known" issue | VMSSPT::JENKINS | Kevin M Jenkins VMS Support Engineering | Wed May 28 1997 09:15 | 15 |
|
You may go ahead and elevate this, but be prepared. We've been around
the block on this already... Basically it has been stated that no
change can be made to DKDRIVERs timeout mechanism. Hence, the best you
can do is set SHADOW_MBR_TMO real small and that will allow a member
of a set to be removed in a couple of minutes.... some playing is
needed to see exactly how small you can set it.
On the other hand, we have been investigating a change to the SHDRIVER
that "might" allow us to "override" the device driver and remove a
member in a specified time, even if we have not heard back from our
Pack-Ack. However at this point this sort of change has some
problems that need to be worked out...
Kevin
|
| 648.4 | When the new SHDRIVER is available ? | AEOENG::ANTHOINE_J | | Fri May 30 1997 04:24 | 10 |
| The SHADOW_MBR_TMO is set at 5 seconds but it has no influence on
the behavior of the DKDRIVER. We have tested the system with different
value from 1 to 60 seconds. The customer would want to detect a disk
fault ASAP, and more precisely in a maximum of 20 seconds.
I will be very happy to test the new SHDRIVER. When is it available ?
Thanks a lot,
Jacques
|