Re: [foxboro] Recovering Blocks
- From: "Johnson, Alex (Foxboro)" <ajohnson@xxxxxxxxxxx>
- To: foxboro@xxxxxxxxxxxxx
- Date: Fri, 31 Oct 2003 13:47:32 -0500
I still think this problem is caused more often by restoring backups than by
what you experienced. :)
I've got to admit that I don't know understand how the problem would have
occurred. You really should write it up and turn in a CAR so that we have an
official record.
Now, on to correcting the problem. The official procedure is (as you know):
1) Upload
2) Save all
3) Initialize
4) Reboot
5) Load all
Or as you suggested upload the compound, save the compound, delete the
compound, reload the compound.
However, there is an unofficial procedure that can be used AT YOUR OWN RISK.
Our you can bring in FS and do it at our risk. :) Be ready to practice,
practice, practice before doing this with a real system. Still, it is
helpful.
Make a SAVE ALL on the on-line machine after an UPLOAD.
Load it on an off-line machine into a CP of the same type and letterbug
name.
Delete the blocks that are in the ICC, but not the on-line CP.
Exit the ICC
At this point, you have a set of files on the off-line AW that are
consistent with the configuration of the on-line CP.
Those files are:
The directory - /opt/fox/ciocfg/<cpLbug>
The directories /opt/fox/ciocfg/<cmpdName>
Where <cmpdName> is the name of any compound that has Sequence Blocks (IND,
DEP, EXC, etc) or PLB blocks in it.
Tar them off
Tar them onto the on line machine.
Now the workfile and the CP match.
Use the CSA utilities to bring CSA in-line if necessary - boils down to dump
to ASCII file, edit file, reload.
Bingo, you are consistent without a reboot.
Alex Johnson
Invensys Systems, Inc.
10707 Haddington
Houston, TX 77043
713.722.2859 (office)
713.722.2700 (switchboard)
713.932.0222 (fax)
ajohnson@xxxxxxxxxxx
For the latest information on ArchestrA, go to
www.invensys.com/Archestra.html.
-----Original Message-----
From: William C. Ricker [mailto:wcricker@xxxxxxxxxxxxxxx]
Sent: Friday, October 31, 2003 11:52 AM
To: foxboro@xxxxxxxxxxxxx
Subject: Re: [foxboro] Recovering Blocks
>The most likely cause is that you restored the AW that hosts the CP from a
>backup and that backup did not have all of the blocks.
Ummm... Not nessessarily.
My story (happened just this past week)
On hiting the DONE key for a block edit, I got a pop up saying there was
an error in communicating with the CP. Not good. SYSMON shows a CP with
one
side failed. I exit ICC and the time it takes is consistant with it doing a
checkpoint. Quick trip to the downstairs to the rack room and sure enough,
one side of the CP is red, the other green.
Back upstairs, the operators see a flash of backrounds on screen where all
backrounds turn yellow for 5 or 10 seconds. Look at SYSMON and it shows
both sides of the CP are now ok. Time since initial error message around
10 minutes now.
Everything looks ok, so back to editting. After a bit we notice that a
block
that was put in 15 or 20 minutes before the failure gets a NOT FOUND error
when trying to edit it. Looking around we find that none of the edits made
since the last checkpoint before the failure are in the CP, though they do
appear in ICC.
So, the first question is what exactly happened. It looks as if one side
of the CP (the failed side) reloaded while the other side was checkpointing,
then synced up (we saw no process bump). Is that possible? Thinking about
it I don't know why not. (Oh, and the printers on that node are not up, so
no messages.)
Second is how to recover. We know exactly that all edits were made in a
single compound. So, for this instance only, my advice to the customer
is that the compound should be deleted and restored, either
"delete-undelete"
or save to disk, delete, restore from disk. There is a planned shut of the
process area soon where this might be accomplished.
Alex, your comments on both questions would be welcome
>There are Helpful Hints on what to do
>
>You can also get CSA out of date this way. Restoring it involves generating
>some text files using the CSA utilities, editing them, and reloading CSA..
>There is an HH for this too I think and there is a "manual" for the CSA
>commands.
Would this be in that On-Line Upgrade document that was floating around a
while ago?
Regards,
William C. Ricker
FeedForward, Inc.
Marietta, GA, USA
wcricker@xxxxxxxxxxxxxxx
770-426-4422
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
Other related posts:
