ADSTAR Distributed Storage Manager (ADSM) Server Licensed Materials - Property of IBM (C) Copyright IBM Corporation 1990, 2000. All rights reserved. ------------------------------ ATTENTION -------------------------------- This is a service PATCH based on certified service level 3.7.3.0. ------------------------------------------------------------------------ Service Level This is a smit/installp installable patch that will install as level 3.7.3.6 for the fileset tivoli.tsm.server.rte42. We strongly advise that you install this patch as an UPDATE (set "COMMIT software updates?" to no and "SAVE replaced files?" to yes ------------------------------------------------------------------------ Testing has focused on the following items. ------------------------------------------------------------------------- IC26881 LABEL LIBVOLUME W/ OVERWRITE=YES CAN OVERWRITE THE LABELS OF VOLUMES THAT ARE PART OF A STORAGE POOL AND HAVE DATA. The ADSM/TSM server may re-label a volume defined to a storage pool or in the volume history file under the following conditions: 1) The user specified the OVERWRITE=YES option in the LABEL LIBVOLUME command. 2) The volume defined to the storage pool or in the volume history file was either checked out or moved to a previously empty slot within the library. <@> IC26903 ADSM DISMOUNT FAILURE RETRY LOGIC IN EMM ENVIRONMENT HALTS SERVER In an External Media Manager environment ADSM/TSM should not be using a dismount failure retry logic. Errors messages that maybe present during failure: ANR9999D mmsext.c(2822): Error writing 'DISMOUNT LIB1 H00612 ' to exit. (-1/21) ANR9999D mmsext.c(691): unable to close file, rc = -1. ANR8769E External media management function DISMOUNT returned result=**Undefined**. <@> IC26979 TSM 3.7.3 DOES NOT POLL DRIVES INDEFINITELY WHEN 3494SHARED YES OPTION IS SPECIFIED/THE DRIVE WILL BE POLLED FOR 10MIN AND OFF-L The TSM 3.7.3 server does not poll drives indefinitly when the 3494SHARED YES option is specified in the dsmserv.opt file and there is another application's volume loaded in the drive. . TSM polls the drive for 10 minutes and then marks the drive off-line. <@> IC26995 TSM 3.7.3.0 SERVER DOES NOT ALLOW THE USER TO UPDATE OR DELETE AFTER THE DRIVE HAS BEEN MARKED OFF-LINE. The TSM 3.7.3.0 Server will poll a shared tape drive if it believes that there is a cartridge loaded in the specified drive that is in use by another application. If the TSM Server marks the drive off-line, after polling the drive for a period of time, the user cannot update or delete the drive until the TSM Server is recycled <@> IC27060 UPGRADING THE TSM SERVER TO 3.7.3.0 MAY CAUSE THE TSM SERVER TO HANG WHILE MOUNTING THE RECOVERY LOGS. TSM WILL NOT START. When the server is started, the consistency checks performed on the log are incorrect and may lead to a hang condition or erroneous reports of log corruption, resulting in the log being needlessly truncated. <@> IC27077 ISSUING THE UPDATE DRIVE COMMAND WITHOUT SPECIFYING THE DEVICE NAME IN THE COMMAND COULD CAUSE 3494 TO LOSE DRIVE INFORMATION After an UPDATE DRIVE command without the DEVICE parameter, the TSM server looses the device number for the given drive. The TSM server needs that device number to specify the drive when communicating to the 3494 Library Manager. The TSM server may abend when attempting to communicate to the 3494 Library Manager to perform a mount/demount activity or to obtain information about the given drive. Otherwise, there may be I/O errors in the activity log; that is, ANR8301E messages with a completion code of 0x023. <@> IY08870 ANR9999D AFMIGR.C(2805): ERROR DELETING TRANSACTION FOR 0 BYTES MOVED RECLAMATION, PROCESSING COPY STORAGE POOL VIRTUAL VOLUME The problem occurs when both on and off site reclaimation are running at the sametime and both need the same volume. When this occurs one of the reclaimations will fail with an error reporting zero bytes moved, and on the Sun and NT platform the server will abort. <@> ********************************************************************** * MOVE DRMEDIA command * ********************************************************************** The following value was added to the REMOVE parameter of the MOVE DRMEDIA command to specify that volumes should be ejected from a SCSI library if there are available slots in the entry/exit port: o REMOVE=Untileefull Command Syntax: >>--MOVe DRmedia--volume_name------------------------------> +---REMove-=-Yes------------+ | | >----+---------------------------+------------------------>< | | +--REMove-=-+--No-----------+ | | +--Yes----------+ | | +--Bulk---------+ | | +--Untileefull--+ >-----------------.......----------------------------------> REMOVE=Untileefull Specifies that the volume is ejected from the library if there is at least one empty slot in the entry/exit port. If the volume is ejected from the library, the volume entry is deleted from the TSM library inventory. This option only applies to a SCSI library. If the entry/exit port is full, the MOVE DRMEDIA command fails and any remaining eligible volumes are NOT processed. After removing the volume(s) from the entry/exit port, the MOVE DRMEDIA command must be re-issued to process any remaining eligible volumes. The REMOVE=UNTILEEFULL parameter behaves like the REMOVE=YES parameter until the entry/exit port is full. If there is an available slot in the entry/exit port, the MOVE DRMEDIA command moves the volume to the next available slot in the entry/exit port, issues a message to indicate that the volume should be removed from the entry/exit port, and waits until a 'REPLY id' is issued. <@> PQ38261 FIXEDIOBUFSIZE DATA-INTEGRITY ERRORS PRIMARY STORAGE POOLS This problem will only occur on an MVS server when the user explicitly specifies FIXEDIOBUFSIZE greater than 0. When the error is detected, the following messages may be found in the JES Job log during space reclamation, audit volume, export node, retrieves or restores ANR9999D SSRECONS(2212): Invalid magic number found in frame header. ANR9999D SSRECONS(2219): Wrong segment group id found in fram header. ANR9999D SSRECONS(2225): Wrong sequence number found in frame header. ANR9999D SSRECONS(2231): Actual: Magic=00000000, SrvId=0, SegGroupId=0, ANR9999D SeqNum=0. ANR9999D SSRECONS(2238): Expected: Magic=E2C7C6D9, SrvId=0, SegGroupId=7438494- ANR9999D 3. SeqNum=4. or ANR1165E Data-integrity error detected for file in storage pool: Node, Type Backup, File space:, File name ANR9999D SMNQR(1286): Integrity error detected for bitfile. <@> AUDIT VOLUME Enhancements (A result of APAR PQ38261) ------------------------- The AUDIT VOLUME command has been enhanced so that it is more effective in detecting problem files. If an inconsistent file is detected, message ANR2308W, ANR2316W, or ANR2317W is issued. These messages have been enhanced to include the version number of the file and the total number of versions. The versions are numbered from most recent (1) to the least recent (n, where n is the total number of versions). The action taken by the server when a problem file is detected has not been changed. NOTE: If corrupted files are deleted during a volume audit with FIX=YES, a subsequent audit of the same files may produce messages indicating invalid frame headers, without listing any damaged or deleted files. This scenario occurs only in certain situations, usually involving imported files. Should this occur, do not be concerned. The frame-header errors indicate that an aggregate still contains structural damage, even though any corrupted files have already been removed. During the next reclamation of a volume containing this data, the aggregate will be reconstructed and a later audit will not report a frame-header problem.