06-Aug-2011

Maintenance
This month’s maintenance-run yieled no problems – except that the next morning, there was a list of logfiles but none was available for review. A known issue – matter of wrong order, unless I decide to keep the files available for a few more days. Anyway – I got the mail stats:

PMAS statistics for July
Total messages    :   3354 = 100.0 o/o
DNS Blacklisted   :    311 =   9.2 o/o (Files: 31)
Relay attempts    :   2439 =  72.7 o/o (Files: 31)
Accepted by PMAS  :    604 =  18.0 o/o (Files: 31)
 Handled by explicit rule
        Rejected :     94 =  15.5 o/o (processed),   2.8 o/o (all)
        Accepted :    228 =  37.7 o/o (processed),   6.7 o/o (all)
 Handled by content
       Discarded :    107 =  17.7 o/o (processed),   3.1 o/o (all)
    Quarantained :    127 =  21.0 o/o (processed),   3.7 o/o (all)
       Delivered :     48 =   7.9 o/o (processed),   1.4 o/o (all)

Once more a rather large amount of relay attempts; the vast majority in just three days: 01-Aug-2011 (a file of 143 blocks), more on 07-Aug-2011 (191 blocks) and a smaller (116 blocks) on 26-Aug-2011. in between smaller files, or none at all.
I found the license of the spam filter would expire 19-aug-2011, so I requested a new VMS Hobbyist license, and with that, a new license for PreciseMail AntiSpam. So I’m protected for ome more year.

Battery gone bad
Scanning the logfile on 04-Aug, i found out that there has been a hick-up in the cluster: the connection to the quorum disk had gone, causing the server to stall for a fe seconds; but the connection was regained, and processing continued:

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:18.95 Node DIANA (csid 00010001) failed to read quorum disk

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:18.95 Node DIANA (csid 00010001) lost “connection” to quorum disk

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:18.95 Node DIANA (csid 00010001) proposed modification of quorum or quorum disk membership

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:18.95 Node DIANA (csid 00010001) completed VMScluster state transition

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:21.95 Node DIANA (csid 00010001) re-established “connection” to quorum disk

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:21.95 Node DIANA (csid 00010001) proposed modification of quorum or quorum disk membership

%%%%%%%%%%% OPCOM 3-AUG-2011 14:42:28.01 %%%%%%%%%%%
14:40:21.95 Node DIANA (csid 00010001) completed VMScluster state transition

But afterwardsm nothing seemed wrong, only the number of errors on almost all disks has gone up to 6 – except for the quorum disk that now has 15 errors.
Not able to access the system at that time, no earlier than this weekend.
I ran diagnose and found out that the battery of my HSZ50-controller was declared bad, which caused write-through cache to be disabled and that caused resets of the bus. All in all 6 times – which explaines the 6 errros on all disks.
Now what: Either obtain a new backup-battery block, at least one with a good battery, or remove the shared SCSI completely. It’s no big deal, just a lot of work, but it can be combined with other work that is currently in progress: Setting up a new WASD configuration and updates of the software involved.
FreeAXP problems
I’ve been using FreAXP – Migration Specialties Inc.’s Alpha-400 emulator – on Windows XP, and on Vista. The first never gave me a problem, but on Vista, it was quite another story. It runs with no problem, just it won’t connect to the loopback adaptor. Even if this is setup with a fixed address that matches the network….
With Windows 7, it seems to run in the 32-bit version, and you cabn even access it over the loopback adaptor. But no matter what configuration, in the 64-bit version it simply failes to conect.
The only thing I didn’t try yet, is install it usinng the Administrator account – that I found “disabled” – by default?