Error %PMDF-W-ERRMAKELOG on VMS
%PMDF-I-VERSION, PMDF version is PMDF V6.3
AlphaServer ES40 running OpenVMS Alpha V8.3
PMDF_SHARE_LIBRARY version V6.3-x11; linked 19:47:06, May 31,2007
Starting today I'm seeing in the operator log:
%%%%%%%%%%% OPCOM 5-MAY-2008 13:04:32.98 %%%%%%%%%%% (from node
VRENT at 5-MAY-2008 13:04:32.95)
Message from user PMDF_SERVICE on VRENT
%PMDF-W-LOGOPENFAIL, Cannot open PMDF log file: %RMS-E-CRE, ACP file
create failed
(Image DSA8:[SYS0.SYSCOMMON.][SYSEXE]MAIL_SERVER.EXE user PMDF_SERVICE
id 2096f5b9)
%%%%%%%%%%% OPCOM 5-MAY-2008 13:04:32.98 %%%%%%%%%%% (from node
VRENT at 5-MAY-2008 13:04:32.95)
Message from user PMDF_SERVICE on VRENT
%PMDF-W-ERRMAKELOG, Error making log entries: %NONAME-W-NOMSG, Message
number 6ADC1CC8
(Image DSA8:[SYS0.SYSCOMMON.][SYSEXE]MAIL_SERVER.EXE user PMDF_SERVICE
id 2096f5b9)
I'm thinking it's a file version > 32676 problem but haven't found
anything like that yet. What is MAIL_SERVER.EXE used for and where does
it want to write a log? I'm using Decnet V.
Mail appears to get sent/received except for two messages to local
mailing lists in the local queue with histories of:
ARTZER: Internal error in VMS MAIL detected.
ASHER: Internal error in VMS MAIL detected.
...
Probably an effect of the log problem.
Any ideas where to look? Suppose I'll enable debugging on the local channel.
It's probably a coincidence that I upgraded one of 2 nodes from 7.3-2 to
8.3 (+ latest patches) on Friday. Both nodes are now at 8.3. One of
those patches was VMS83A_MAILSHR-V0100. This was my first though but the
above error message is coming from both nodes however - not just the one
updated.
--
-- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison
-- karcher-AT-waisman.wisc.edu
RE: Error %PMDF-W-ERRMAKELOG on VMS
It is possible that the INDEXF.SYS has had too many extents, and you can no longer create files on disk 9even if it is not full). I cannot tell if your pmdf_log files are on the system drive. You will not be able to create new files on that drive at all until you either delete files or re-initialize the drive with and an increase /headers qualifier. Deleting files is a temporary workaround in this case.
============================================
Joe Prchlik
Director of Network/System Services
Northwest Ohio Computer Association
E: [email]prchlik@nwoca.org[/email]
P: 419.267.2515
F: 419.267.5248
A: 22-900 State Route 34 Archbold, OH 43502
-----Original Message-----
From: Carl Karcher [mailto:karcher@Waisman.Wisc.Edu]
Sent: Monday, May 05, 2008 2:49 PM
To: [email]info-pmdf@process.com[/email]
Subject: Error %PMDF-W-ERRMAKELOG on VMS
%PMDF-I-VERSION, PMDF version is PMDF V6.3
AlphaServer ES40 running OpenVMS Alpha V8.3
PMDF_SHARE_LIBRARY version V6.3-x11; linked 19:47:06, May 31,2007
Starting today I'm seeing in the operator log:
%%%%%%%%%%% OPCOM 5-MAY-2008 13:04:32.98 %%%%%%%%%%% (from node
VRENT at 5-MAY-2008 13:04:32.95)
Message from user PMDF_SERVICE on VRENT
%PMDF-W-LOGOPENFAIL, Cannot open PMDF log file: %RMS-E-CRE, ACP file
create failed
(Image DSA8:[SYS0.SYSCOMMON.][SYSEXE]MAIL_SERVER.EXE user PMDF_SERVICE
id 2096f5b9)
%%%%%%%%%%% OPCOM 5-MAY-2008 13:04:32.98 %%%%%%%%%%% (from node
VRENT at 5-MAY-2008 13:04:32.95)
Message from user PMDF_SERVICE on VRENT
%PMDF-W-ERRMAKELOG, Error making log entries: %NONAME-W-NOMSG, Message
number 6ADC1CC8
(Image DSA8:[SYS0.SYSCOMMON.][SYSEXE]MAIL_SERVER.EXE user PMDF_SERVICE
id 2096f5b9)
I'm thinking it's a file version > 32676 problem but haven't found
anything like that yet. What is MAIL_SERVER.EXE used for and where does
it want to write a log? I'm using Decnet V.
Mail appears to get sent/received except for two messages to local
mailing lists in the local queue with histories of:
ARTZER: Internal error in VMS MAIL detected.
ASHER: Internal error in VMS MAIL detected.
...
Probably an effect of the log problem.
Any ideas where to look? Suppose I'll enable debugging on the local channel.
It's probably a coincidence that I upgraded one of 2 nodes from 7.3-2 to
8.3 (+ latest patches) on Friday. Both nodes are now at 8.3. One of
those patches was VMS83A_MAILSHR-V0100. This was my first though but the
above error message is coming from both nodes however - not just the one
updated.
--
-- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison
-- karcher-AT-waisman.wisc.edu
Re: Error %PMDF-W-ERRMAKELOG on VMS
Joe Prchlik wrote:[color=blue]
> It is possible that the INDEXF.SYS has had too many extents, and you can
> no longer create files on disk 9even if it is not full).[/color]
After a bit more poking this error is merely an effect. The real error
was in the task_server*.log files made by the process symbiont:
%MAIL-E-OPENOUT, error opening PMDF_QUEUE:[000000]MAIL.TXT; as output
-RMS-E-CRE, ACP file create failed
-SYSTEM-F-EXQUOTA, process quota exceeded
%MAIL-E-OPENOUT, error opening PMDF_QUEUE:[000000]MAIL.TXT; as output
-RMS-E-CRE, ACP file create failed
-SYSTEM-F-EXQUOTA, process quota exceeded
I ran the local (l) queue manually (under my own account quotas) and it
processed the two messages stuck there just fine. These messages had a
very long to: list (750+ local addresses) which grows weekly so it looks
like a process quota is being exhausted when a post to this list gets
delivered. The process symbiont appears to get it's quotas from the
PMDF_SERVICE account which 'look' reasonable:
Maxjobs: 0 Fillm: 300 Bytlm: 64000
Maxacctjobs: 0 Shrfillm: 0 Pbytlm: 0
Maxdetach: 0 BIOlm: 150 JTquota: 8192
Prclm: 10 DIOlm: 150 WSdef: 8192
Prio: 3 ASTlm: 250 WSquo: 28672
Queprio: 0 TQElm: 40 WSextent: 65536
CPU: (none) Enqlm: 2048 Pgflquo: 400000
Accounting log shows a termination message of:
Final status text: %SYSTEM-F-EXITFORCED, forced exit of image or process
by SYS$
Page faults: 7901 Direct IO: 648
Page fault reads: 792 Buffered IO: 1106
Peak working set: 11968 Volumes mounted: 0
Peak page file: 201792 Images executed: 32
Hard to tell which quota but it's not Pgflquo. I'd bet on Bytlm. Looks
like figuring out which quota is going to be a real hunt and perhaps a
job for availman. Of course, I'll have to wait for the next post to the
big mailing list as I don't have any more test data.
--
-- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison
-- karcher-AT-waisman.wisc.edu
