> Hi,
> I know this is an ADSM list, but this is the only place where I
> can find so many 'experts' in one place.
As you probably know, WDSF support went away in March. But, don't feel
like you have no recourse. You do. I have been through the abend you
mention more times than I care to admit (more than 10 times). Consider
yourself VERY fortunate that you have a small server.
First, did you try an audit?
Try: DSSERV DBAUDIT FIX VERBOSE
>
> I have a small WDSF server with a dozen or so OS/2 clients and 5 AIX
> clients - like I said, small! All works fine except for one of the
> OS/2 clients. Here is a section of the console log:
>
> Server startup and a few sessions deletedY
>
> 04/28/1994 15:05:39 ANS0100I Session 5 started for node CMIP-DC (OS/2).
> 04/28/1994 15:06:31 ANS0120W Protocol error on session 5 - invalid verb
> header received.
> 04/28/1994 15:06:32 ANS0117W Session 5 terminated - protocol violation
> detected.
>
> A few more normal sessions deletedY
>
> 04/28/1994 23:26:01 ANS0100I Session 14 started for node CMIP-DC (OS/2).
>
>
> ANS0001S Server ABORTING - internal error SMU009 detected.
F0
Ahhhh, brings cold sweats to my body............
If it still abends, you will need to dump and restore your database.
The following procedure works quite well and should not fail. Please
let me know if you have any problems. I consider myself an expert in
this function (kinda like being an expert in having teeth pulled :-))
Best of luck....
To: msako AT lsil DOT com
Content-Length: 4094
Status: RO
X-Lines: 100
Instructions for performing a DUMP and RESTORE of a WDSF/VM database.
========================================================================
If you can currently start your server issue the following commands to
If you can currently start your server issue the following commands to
flush any buffered data to disk:
TRACE INSTRON
CKPT
CKPT
CKPT
TRACE INSTROFF
Before you start, backup your WDSF/VM system. Make sure that the
server has been HALTed before performing this backup.
Follow these steps (1-5) if you want to use DASD as your work area.
(If you would like to use TAPE as your work area, please
see the instructions below.)
1. You'll need a workdisk (can be temporary) large enough to hold
all of your current database extents. This minidisk will be used
to dump the database to during the DUMP portion of the
DSSERV DUMP/RESTORE.
2. Enter the following commands :
ACCESS xxx B (xxx is the address of the workdisk )
FILEDEF CONV DISK TEST DUMP B (RECFM F LRECL 32760 BLKSIZE 32760
DSSERV DBDUMP CONV
3. Run the DSINST command to re-prepare your system. This command
is documented in the WDSF/VM Administration Guide.
IMPORTANT NOTE!!!!: Make sure you enter 0 when prompted for the
number of repository disks to be prepared!! Any value greater
than 0 will cause your repository to be destroyed!!
4. Enter the following commands:
ACCESS ddd B (The ddd should be the address of your workdisk)
FILEDEF CONV DISK TEST DUMP B (RECFM F LRECL 32760 BLKSIZE 32760
DSSERV DBREST CONV
5. When this is done, bring up the server and immediately run an
AUDIT by issuing the following command to bring up the server:
DSSERV DBAUDIT FIX VERBOSE
Follow these steps (1-5) if you want to use TAPE as your work area.
1. You will be performing the dump/restore using tapes that contain
standard labels, so make sure each tape contains a standard
label. Standard labels can be written by issuing the following
command:
TAPE WVOL1 <volid> ( SL
where: volid - is the volume id of the tape you are using.
2. Enter the following commands :
LABELDEF <ddname> VOLID SCRATCH
FILEDEF <ddname> TAP1 SL ( RECFM F LRECL 32760 BLKSIZE 32760
DSSERV DBDUMP <ddname>
where: ddname - is a 1-6 character DD name.
NOTE: What happens when each tape fills up depends on the way
your system is set up. By default a routine named
DMSTVS will gain control and handle the EOV processing.
This routine will send a message to the DSSERV console
and to the operator console, asking for a tape. If
this happens you will be given 3 choices and you should
respond with choice 3 (NEWTAPE). If, however, your
system contains the user exit module DMSTVI (not shipped
with CMS) then this module will gain control and handle
the EOV processing. This module is a user exit that
gives each installation the ability to define the way
tape processing should be performed. If you have done
this on your system then you need to make sure this
module will prompt for a new tape, or you will need to
rename it so that it will not be invoked.
3. Run the DSINST command to re-prepare your system . This command
is documented in the WDSF/VM Administration Guide.
IMPORTANT NOTE!!!!: Make sure you enter 0 when prompted for the
number of repository disks to be prepared!! Any value greater
than 0 will cause your repository to be destroyed!!
4. Enter the following commands:
LABELDEF <ddname> VOLID SCRATCH
FILEDEF <ddname> TAP1 SL ( RECFM F LRECL 32760 BLKSIZE 32760
DSSERV DBREST <ddname>
where: ddname - is a 1-6 character DD name.
5. When this is done, bring up the server and immediately begin
an AUDIT by issuing the following:
DSSERV DBAUDIT FIX VERBOSE
After the audit completes, the server should operate normally.
________________________________________________________________________
| Mitch Sako \\\\ \\\\ \\\\\\\\ |||||||| /////// // // |
| I-NET Corporation \\ \\ \\ \\ \\ || // // // |
| LSI Logic Contract \\ \\\ \\ \\ || // //////// |
| Mailstop E-195 \\ \\ \\ || // // // |
| 1501 McCarthy Blvd. \\ \\ \\\\\\\\ || /////// // // |
| Milpitas CA 95035 |
| local: mitch@asic Phone: (408) 433-4187 |
| internet: msako AT lsil DOT com FAX: (408) 433-8796 |
| uucp: lsil!msako Pager: (408) 989-3365 |
| ibmmail: USMILUN9 AT IBMMAIL |
| DISCLAIMER: Opinions expressed are mine alone |
|______________________________________________________________________|
|