[an error occurred while processing this directive]
HP OpenVMS Systems |
HP Advanced Server V7.3B for OpenVMS
|
Previous | Contents | Index |
This chapter describes problems resolved by software changes made in
Advanced Server V7.3A-ECO1 for OpenVMS. These software changes will also be included in future
ECOs.
8.1 Advanced Server Problems with File Access Permissions
Problem:
http://www.microsoft.com/technet/prodtechnol/windows2000serv /reskit/distsys/dsintro.asp |
These three problems may show up in any of a variety of unexpected ways regarding user access to file and directory resources. Some problems that have been reported include users being able to create files but not being able to modify or rename them, and users being unable to access a file that they expect to be accessible. For instance, when you display the permissions for the problem directories, as in the following example (using the ADMINISTRATOR SHOW FILES/PERMISSIONS command), the permissions might include "Special Access" and "(Not Specified)" instead of the expected permissions (such as "Change (RWXD) (RWXD)").
ACCNTS <DIR> Permissions: Access Administrators Special Access (All) (Not Specified) Everyone Special Access (RWXD) (Not Specified) DWilson Special Access (All) (Not Specified) Server Operators Special Access (RWXD) (Not Specified) SYSTEM Special Access (All) (Not Specified) |
Note also that the CREATOR OWNER information is missing from this display. The following example shows the permissions that might be expected for the directory ACCNTS (and here the CREATOR OWNER information is present, as it should be):
ACCNTS <DIR> Permissions: Access Administrators Full Control (All) (All) CREATOR OWNER Full Control (All) (All) Everyone Change (RWXD) (RWXD) DWilson Special Access (All)* (Not Specified) Server Operators Change (RWXD) (RWXD) |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
8.2 Wrong Time Set on Client After NET TIME Command Is Used to Synchronize with the Server
Problem:
When the NET TIME command is issued from a PC client to synchronize its clock with the Advanced Server (V7.3-ECO2 or later), as in the following example where the server is named TINMAN, the time set on the client might be incorrect. The time set on the client will be off by the GMT difference (in hours) reported by the server.
C:\> NET TIME \\TINMAN /SET |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
8.3 File Remains Locked, Access Violation Errors Reported
Problem:
In rare circumstances, usually seen in environments that include a large number of clients, an attempt to open a file in a share on an OpenVMS Cluster can fail with a sharing violation. This continues to occur after repeated attempts to open the file.
The problem is seen only on one server node in the cluster. The file can be opened from the other server node(s) in the cluster. The only way to solve the problem is to restart the server on the node that has the problem.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
8.4 Copying New Files with Certain Names to a Share Might Fail
Problem:
Copying new files to an Advanced Server share might fail, depending on the names of the files. The following are examples of file names that cause the failure (the file extension is irrelevant):
00000001.*
00000010.*
00000011.*
00000100.*
00000101.*
00000110.*
00000111.*
Attempts to copy such files result in an error message similar to the following:
Error Copying File Cannot copy filename: Cannot find the specified file. Make sure you specify the correct path and filename. |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
8.5 Server Might Crash in Routine RealGetSidOfUnixOwner
Problem:
The server might crash in routine NtaclGetSidOfUnixOwner of module NTACLMAP, leaving a traceback similar to the following in the PWRK$LMSRV log file:
image module routine line rel PC abs PC %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=0000000000000000, PC=00000000001AC7E0, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV ACLMAP RealGetSidOfUnixOwner 120640 0000000000000250 0000000000... PWRK$LMAPISHR NTACLMAP NtaclGetSidOfUnixOwner 89523 00000000000001FC 00000000005... PWRK$LMAPISHR NTACL get_custom_sd 6385 00000000000000B0 00000000005C... PWRK$LMAPISHR NTACL NtaclFetchWorkingSD 98211 00000000000016D8 00000000005... . . . |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
8.6 PWRK$LMMCP Hangs with One or More Server Cluster Members in the LEF State
Problem:
On an OpenVMS Cluster of four or more cluster members, an attempt to start the Advanced Server on the 4th and remaining cluster members might result in the PWRK$LMMCP process on each of these servers hanging in the LEF (local event flag) wait state. Symptoms include clients being unable to connect to the servers and ADMINISTER commands hanging or timing out.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO1 for OpenVMS.
This chapter describes problems resolved by software changes made in
Advanced Server V7.3A-ECO2 for OpenVMS. These software changes will also be included in future
ECOs.
9.1 Execution of ADMINISTER SHOW Commands Causes Server to Crash with Access Violation in PWRK$LMRPCSHR
Problem:
In rare circumstances when using ADMINISTER SHOW commands, the server might crash with an access violation in PWRK$LMRPCSHR. Error messages and a traceback similar to the following would be logged:
SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address= 0000000000000001, PC=FFFFFFFF80D8A8E8, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC 0 0000000000000000 FFFFFFFF80D8A8E8 PWRK$LMRPCSHR DYNAMIC add_file_to_list 30902 00000000000002D8 0000000000427048 PWRK$LMRPCSHR DYNAMIC library_load 30939 00000000000004C0 0000000000427230 PWRK$LMRPCSHR RPCXPORT rpc__xport_info 27506 00000000000000BC 0000000000437B8C PWRK$LMRPCSHR RPCBIND rpc__binding_alloc 30784 0000000000000088 0000000000433B08 . . . |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.2 PWRK$LMSRV Crash While Enumerating Shares
Problem:
A client attempting to enumerate the shares on a server causes a PWRK$LMSRV crash in NetrShareEnum on address 0x00000004
%SYSTEM-F-ACCVIO, access violation, reason mask=04, virtual address=0000000000000004, PC=00000000000BA448, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV ENUMS NetrShareEnum 81872 0000000000000D48 00000000000BA448 PWRK$LMSRV SRVSVC_S srvsvc_NetrShareEnum 22687 0000000000004418 00000000009C398 |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.3 PWRK$KNBDAEMON and PWRK$NBDAEMON Log Files Cannot be Viewed While Advanced Server is Running
Problem:
When attempting to troubleshoot TCP/IP connectivity problems, it is impossible to get any useful information from the PWRK$LOGS:PWRK$KNBDAEMON_nodename.LOG file. You cannot even enter the $TYPE or $EDIT command. The file is locked by the PWRK$KNBDAEMON process, and its buffers never seem to get flushed until the file is closed when the process exits. Thus, you will have to stop Advanced Server to view the files. This file is the ONLY source for errors when the server has run out of (TCP/IP or streams) sessions. However, it is nearly impossible to tell that is truly the issue while the server is running.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS. It is now possible to view
the log files for KNBDAEMON and NBDAEMON process, using the $TYPE or
$EDIT command. You do not have to shut down Advanced Server to view
PWRK$KNBDAEMON and PWRK$NBDAEMON log files.
9.4 Server Announcements to Downlevel Servers Not Qualified
Problem:
In previous versions of the Advanced Server, the Advanced Server software sent a server announcement to the <domain> 00 name, which is the name on which Microsoft LAN Manager servers receive server announcements. The Advanced Server software was supposed to send this announcement only if the registry value lmannounce located in the key HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LanmanServer\Parame ters was set to 1 but Advanced Server was ignoring this value.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS. Advanced Server now sends
the server announcements only if lmannounce is 1. Lmannounce is 0 by
default.
9.5 Cannot Create a Share to an ISO-9660 CD-ROM Device
Problem:
When the Server is configured for a language that does not map to the server character set ISO 8859-1, an ISO 9660 CDROM drive cannot be shared nor accessed. The following error is returned if you attempt to create a share to a mounted ISO 9660 CDROM disk.
KSGILE.DOM\\KSGILE> add share cdrom nt40srv_en$:[000000] %PWRK-E-ERRADDSHARE, error adding share "CDROM" -LM-E-NERR_UNKNOWNDEV, the device or directory does not exist |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.6 Server Might Crash in Routine LsarLookupNames
Problem:
The server might crash with an access violation (ACCVIO) in module LSARPC, routine LsarLookupNames. A trace back similar to the following would be seen in the PWRK$LMSRV log file:
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=0000000000000000, PC=00000000000E8CE8, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV LSARPC LsarLookupNames 88655 0000000000000938 00000000000E8CE8 PWRK$LMSRV LSARPC_S lsarpc_LsarLookupNames 28720 00000000000033A4 00000000000DDB94 PWRK$LMRPCSHR RPCCALL rpc__call_put 28450 0000000000000240 0000000000686F60 PWRK$LMRPCSHR CNSRV rpc__ncacn_process_request 33181 0000000000001F00 0000000000692830 PWRK$LMRPCSHR CNSRV rpc__ncacn_datawork 33504 0000000000002770 00000000006930A0 PWRK$LMRPCSHR RPCCALL callwork 28523 0000000000000500 0000000000687220 |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.7 The ADMINISTER Command-line Interface Cannot be Used to Manage the Member Server
Problem:
If the current server being managed by ADMINISTER Command (either a local or remote domain/server) that contain a domain name and server name that cannot be displayed with the maximum prompt length of 32 characters. The ADMINISTER Command utility continuously displays the following error:
Error %CLI-E-STRTOOLNG, string argument is too long - shorten |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS. The ADMINISTER Command utility will now truncate the domain name with "..." when required.
9.8 "Cannot Convert Unicode to the Target Server Character Set" Messages in Common Event Log
Problem:
Administer/Analyze (which displays the events in the Advanced Server common event log) might report a large number of the following events:
Event Time : 27-MAY-2002 16:24:32.30 Node: MYNODE Process Id : 210036C6 Event : LAN Manager server error Event Source: LAN Manager Server Event Class : Error Text: Cannot convert Unicode to the target Server character set. |
These messages are posted when a client sends data (usually a file path) to the server that cannot be converted to the server's currently configured language. These messages fill up the event log and are not required to manage the currently running server. These events are no longer posted.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.9 PWCONVERT Cannot Convert SPACE (__20) Correctly
Problem:
After installing Version 7.3A of the Advanced Server, attempts to use PWCONVERT to convert an ODS-2 file name with an escaped space ('__20') with PWCONVERT into an ODS-5 file name will result in a '^^_' character sequence. This should have been '^_' , which represents a space on ODS-5.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.10 PWRK$RENAME.EXE Image was Not Included in the Distribution Kit
Problem:
Advanced Server Version 7.3A cannot open the shares and files created by Advanced Server Version 7.2 for OpenVMS.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.11 Server Crashes in Routine PFS_releaseuser During ULOGOFF
Problem:
Under rare circumstances the server may crash with the following footprint:
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=FFFFFFFFABCDABC9, PC=FFFFFFFF80830114, PS=00000018 %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC . . PWRK$CSSHR_V7 PFS$LIB PFSReleaseUser 30162 0000000000001B24 00000000004A76E4 PWRK$CSSHR_V7 PFS$OPS PFS_freeuser 28944 000000000000B4B8 000000000049A198 PWRK$LMSRV SESSION user_sess_cleanup 65727 0000000000001878 0000000000121E48 PWRK$LMSRV SESSION Invalidate 65668 0000000000001758 0000000000121D28 PWRK$LMSRV ULOGOFF smbulogoff 49217 000000000000010C 0000000000141CCC |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.12 Access to Data Denied
Problem:
Lotus Notes on a Windows 2000 client would receive the error message "Access to data denied" when the data files are located on the Advanced Server.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.13 System Crashes in PWRK$STREAMSSHR
Problem:
In rare circumstances, OpenVMS may bugcheck with an INVEXCPTN at PWRK$STREAMSSHR_V7+11984, with PWRK$LMSRV being the current process.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.14 Server Crashes in Routine PutSidInCache
Problem:
The server might crash with an access violation (ACCVIO) in module SIDCACHE, routine PutSidInCache. A traceback similar to the following would be seen in the PWRK$LMSRV log file:
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=0000000000000000, PC=00000000002544C0, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV SIDCACHE PutSidInCache 101618 0000000000000410 00000000002544C0 PWRK$LMSRV LOOKUP LsapDbLookupSidsInPrimaryDomain 111879 0000000000002D78 000000000024CCF8 PWRK$LMSRV LOOKUP PegSamrLookupSids 108196 00000000000003BC 000000000024A33C PWRK$LMSRV LSARPC LsarLookupSids 88757 0000000000000B4C 00000000000E8EFC <rpc calls...> |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.15 System Crash Caused by PCIDRIVER
Problem:
Sometimes OpenVMS crashes with a footprint of UNXSIGNAL crash at NSA$DEREFERENCE_RIGHTS_CHAIN+14. This problem was caused by the PCIDRIVER.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.16 The Server Fails to Start On a System with 32GB of RAM
Problem:
The configuration manager's NPAGEDYN calculation, performed during server startup or by ADMIN/CONFIG, can fail on extremely large memory systems.
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.17 PWRK$LMSRV Process Crashes If the Server Uses a User Name Exceeding 20 Characters in Length
Problem:
If a client PC attempts to set up a session with the server using a user name that is greater than 20 characters the PWRK$LMSRV process could crash with the following footprint:
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=000000000A7 %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV SMBWORK smbwork 90291 00000000000018F4 00000000000FA004 PWRK$CSSHR_V7 MTS MTS$CONTEXT_SWITCH 6673 0000000000000104 000000000038B694 PWRK$CSSHR_V7 MTS MTS$MAIN_SCHEDULER ? ? PWRK$LMSRV MTS$MAIN main 5678 00000000000000D8 0000000000091A28 PWRK$LMSRV MTS$MAIN __main 0 0000000000000070 00000000000919C0 0 FFFFFFFF802613F4 FFFFFFFF802613F4 |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.18 Advanced Server Crashes with ACCVIO in Routine ssignon_oneway
Problem:
The Advanced Server might crash due to an ACCVIO at routine ssignon_oneway. A traceback similar to the following would be seen in the PWRK$LMSRV log file:
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=000000000000 02F3, PC=FFFFFFFF81030114, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module rout line rel PC abs PC 0 FFFFFFFF81030114 FFFFFFFF81030114 0 FFFFFFFF811D72F4 FFFFFFFF811D72F4 PWRK$LMAPISHR MALDEBUG lmfree 65126 0000000000000408 000000000053C1D8 PWRK$LMSRV SSIGNON_PROC ssignon_oneway 142475 000000000000015C 00000000001982CC PWRK$LMSRV SSIGNON_PROC ssignon_validateuserv2 142942 0000000000000BA8 0000000000198D18 PWRK$LMSRV SSIGNON_PROC ssignon_processmessage |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
9.19 Pathworks Crashes in Routine get_latest_logoff
Problem:
The Advanced Server Version 7.3A PWRK$LMSRV process crashes with the following footprint, if "Force user logoff after logon hours expire" is enabled in the account policy.
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=000000005057 2ECF, PC=000000000016F108, PS=0000001B %TRACE-F-TRACEBACK, symbolic stack dump follows image module routine line rel PC abs PC PWRK$LMSRV FORCE get_latest_logoff 120938 0000000000000978 000000000016F108 PWRK$LMSRV FORCE force_logoff 120721 0000000000000318 000000000016EAA8 PWRK$LMAPISHR CALLOUT callout_task 64907 00000000000003E8 000000000052B638 PWRK$CSSHR_V7 MTS MTS$CONTEXT_SWITCH 6673 0000000000000104 000000000038B694 PWRK$CSSHR_V7 MTS MTS$MAIN_SC ? ? PWRK$LMSRV MTS$MAIN main 5678 00000000000000D8 0000000000091A28 PWRK$LMSRV MTS$MAIN __main 0 0000000000000070 00000000000919C0 0 FFFFFFFF8027959C FFFFFFFF8027959C |
Solution:
This problem is resolved in Advanced Server V7.3A-ECO2 for OpenVMS.
Previous | Next | Contents | Index |