SLS SLSE02028 Storage Library System for OpenVMS V2.8 ECO Summary
NOTE: An OpenVMS saveset or PCSI installation file is stored
on the Internet in a self-expanding compressed file.
The name of the compressed file will be kit_name-dcx_vaxexe
for OpenVMS VAX or kit_name-dcx_axpexe for OpenVMS Alpha.
Once the file is copied to your system, it can be expanded
by typing RUN compressed_file. The resultant file will
be the OpenVMS saveset or PCSI installation file which
can be used to install the ECO.
Copyright (c) Digital Equipment Corporation 1996. All rights reserved.
PRODUCT: Storage Library System for OpenVMS (SLS)
Media and Device Management Services (MDMS)
OP/SYS: OpenVMS VAX
OpenVMS Alpha
COMPONENTS: SLS$TAPMGRRQ.EXE
SLS$SYSBAK.EXE (When SLS is installed on the system)
MRD$SHR.EXE (When a newer version does not already
exist in the SYS$LIBRARY directory)
SOURCE: Digital Equipment Corporation
ECO INFORMATION:
ECO Kit Name: SLSE02028
ECO Kits Superseded by This ECO Kit: SLSE01028
ECO Kit Approximate Size: 6174 Blocks
Kit Applies To: SLS V2.8
MDMS V2.8
OpenVMS VAX V5.5-2 - V6.2
OpenVMS Alpha V6.1 - V6.2
NOTE: This kit should be applied to systems
where either SLS or MDMS has been
installed. MDMS does not include the
backup and restore functionality of
SLS. MDMS can be installed separately to
support these products:
Archive/Backup System (ABS or PAB),
Hierarchical Storage Management (HSM), or
Sequential Media File System (SMF)
System/Cluster Reboot Necessary: No
ECO KIT SUMMARY:
An ECO kit exists for SLS V2.8 on OpenVMS VAX V5.5-2 through V6.2 and
OpenVMS Alpha V6.1 through V6.2. This kit addresses the following
problems:
o Inventory Fails to Update Database
On Alpha processors, the STORAGE INVENTORY JUKEBOX/ADD command
fails to update the MDMS Volume Database (TAPEMAST.DAT). New
volumes in the jukebox are not added to the database nor are the
slot assignment fields updated. This symptom is known to occur for
HSC/HSJ/HSD connected jukebox devices.
o SLS$TAPMGRRQ Process Terminations
- On Alpha processors, the SLS$TAPMGRRQ process may terminate with
an ACCVIO during execution of a STORAGE INVENTORY JUKEBOX
command. There are not any termination messages output to
operator terminals nor to the TAPMGRRQXEQ.LOG file. The
termination status appears only in the system accounting data
file.
o On both VAX and Alpha processors, the SLS$TAPMGRRQ process
terminates during execution of the
STORAGE INVENTORY JUKEBOX/SLOT=(n,"")
command. The OPCOM message contains incorrect addresses:
SLS$TAPMGRRQ, %SYSTEM-F-ACCVIO, access violation, reason
mask=00, virtual address=00000001, PC=00000000, PSL=007C0000
o Import of Magazine to TL82x Jukebox Fails
On Alpha processors, the command
STORAGE IMPORT MAGAZINE/LOCATION=(L1,L2)
fails with the following error message
%SLS-E-BADIMPLOC, bad value for location parameter in
IMPORT MAGAZINE command
even though the location parameter values are within valid slot
range for the TL82x jukebox.
o Import of Cartridge to TL82x or TL81x Jukebox Fails
On Alpha processors, the command STORAGE IMPORT CARTRIDGE fails
with the following error message
%SLS-E-NOJUKSLOTS, the jukebox does not have any more storage slots
available
even though slots are available in the jukebox. Similarly, the
command STORAGE IMPORT CARTRIDGE/SLOT fails with the error message
%SLS-E-NOSUCHSLOT, specified slot does not exist
even though the specified slot is available in the jukebox.
o SLS$TAPMGRRQ Process Loops Following Jukebox Inventory
On Alpha processors, the SLS$TAPMGRRQ process goes into a loop
after executing the STORAGE INVENTORY JUKEBOX command for
direct-connect jukebox devices. The SLS$TAPMGRRQ and SLS$TAPMGRDB
processes exhibit high CPU time and high direct and buffered I/O
during the hang.
o MDMS commands executed concurrently from separate processes on a
node or from separate nodes in a cluster fail with MRD_START_FAIL
or similar MRD status messages. This problem occurs only for
HSJ/HSD/HSC connected robotic devices.
For MDMS-only and SLS installations, this problem may have
triggered the symptoms described in the next problem description.
For SLS installations, this problem may have triggered the symptoms
described in the problem description
(SLS Installations Only) System Backup Job Fails with NONEXPR Status
o The STORAGE UNLOAD command mishandled error status from robotic
devices and output an Opcom request to unload the volume. It
returned to DCL with successful completion even though the volume
has not been unloaded. The volume's record in the database then
contains inconsistent values for the Drive and Jukebox Status fields.
o (SLS Installations Only) SLS system backups incorrectly handled the
volume switch at end-of-tape for DLT loader devices. On VAX, this
problem caused false MRD_SLOT_FULL or MRD_DRIVE_EMPTY status. On
Alpha, it caused symptoms described in next problem description.
NOTE: This ECO resolves this problem for Direct-SCSI connected
loaders. However, we have some reports of volume switch
failures on HSJ/HSD/HSC connected loaders which have not
been reproducible in engineering labs and is still being
investigated.
o (SLS Installations Only) System Backup Job Fails with NONEXPR Status
On Alpha processors, a system backup FILES_n job may fail with
termination status SYSTEM-F-NONEXPR. Accompanying the failure is
an entry in ACCOUNTNG.DAT showing a termination status of ACCVIO
and an entry in ERRLOG.SYS showing a non-fatal bugcheck in
subprocess SYSBAK_pid_n.
o There is a device-name string compare failure which might have caused
non-robotic devices to be handled as jukebox devices during MDMS load
and unload operations.
NOTE: The problems described above seriously impact the usability of
SLS/MDMS software with HSC/HSJ/HSD connected devices where
concurrent MDMS operations are performed from more than one
process and/or more than one node in a VMScluster configuration.
The problems also impact jukebox devices especially those having
large cartridge libraries such as the TL8xx family of devices
where the STORAGE INVENTORY and STORAGE IMPORT commands are often
used.
INSTALLATION NOTES:
The system/cluster does not need to be rebooted after this kit is
installed.
However, after installing this kit, it is necessary to restart SLS for
all of the changes to take effect.
For a mixed-architecture configuration or other heterogeneous VMScluster
configurations, the kit must be installed on each node having its own
SLS$ROOT directory structure.
This patch can be found at any of these sites:
Colorado Site
Georgia Site
European Site
Files on this server are as follows:
slse02028.README
slse02028.CHKSUM
slse02028.CVRLET_TXT
slse02028.a-dcx_axpexe
slse02028.a-dcx_vaxexe
