SYSMGR in the attic

Again, no real problems
On the system front, nothing weird – once more no weird results in email:

PMAS statistics for July
Total messages    :   1970 = 100.0 o/o
DNS Blacklisted   :     84 =   4.2 o/o (Files: 27)
Relay attempts    :    148 =   7.5 o/o (Files: 29)
Accepted by PMAS  :   1738 =  88.2 o/o (Files: 30)
  Handled by explicit rule
         Rejected :   1046 =  60.1 o/o (processed),  53.0 o/o (all)
         Accepted :    194 =  11.1 o/o (processed),   9.8 o/o (all)
  Handled by content
        Discarded :    311 =  17.8 o/o (processed),  15.7 o/o (all)
     Quarantained :     89 =   5.1 o/o (processed),   4.5 o/o (all)
        Delivered :     98 =   5.6 o/o (processed),   4.9 o/o (all)

but on 1st, 6th and 7th of July, logs on relay attempts used 800 blocks on disk but it seems EOF was < 4 blocks - so the files were deleted. On the hardware front however, changes have been made. Mid July, I found that one of the RAID disks in Diana was bad, which may have contributed to performance issues. It also made a lot of noise I found the right type (MAW300NP: 300 Gb, 10K rpm U320 SCSI with 64-bit connector) via Internet, but in the UK, I ordered 2 and had to pay taxes on import....; placed one in Diana and found that the old 9GB SCSI-2 disk didn't work either so that one was removed as well. Plus that the cause of the bad cooling was found tobe cause by accumulated dust in front. Cleaned that, and after that, Diana was put back into action and is running ever since. During that time without disks, I used Dione - the second DS10 - as a backup for all activities except email because I could not handle mail that was on Diane. Dione could be started as well, even using the (assumed broken) cache memory; found that the batteries are bad so caching is now disabled. After the system had hung - again - and was restarted, I managed to make a backup of all four logical disks, hoping it didn't hang again. But it didn't, and Dione was kept alive for one more day, until Diana was re-activated - now with a new RAID disk and a cleaned front so the fans don't have to run at high speed anymore - CPU temp is now well below 40, without extra help. The last few says we have been plagued by power outages, some less than a minute, one for several hours. But since I do no longer have 7 disks that need to spin up - causing a power surge causing loss of power in the attic - there is no need to gradually startup the systems up here. Some outages - in the night - went unnoticed as long as you didn't look at the date and time of operator logs...

Once again: All t be same (almost)
During our holiday trip, the site became unresponsive for some reason and has been off-line for about 10 days. Nothing worked, the only thing on the LAN I could reach was the router. From there, all traffic was impossible. So the problem was either the router, a cable, or a switch in the attic. As it turned out, it indeed was a cable – from the switch to Diana, so NONE of the connected systems could be reached because DHCP nor DNS could be contacted. Once the connection was restored, all worked again. No restart of any system required….
That is why there are less email messages reported:

PMAS statistics for June
Total messages    :   1593 = 100.0 o/o
DNS Blacklisted   :     42 =   2.6 o/o (Files: 14)
Relay attempts    :    129 =   8.0 o/o (Files: 21)
Accepted by PMAS  :   1422 =  89.2 o/o (Files: 21)
  Handled by explicit rule
         Rejected :    869 =  61.1 o/o (processed),  54.5 o/o (all)
         Accepted :    160 =  11.2 o/o (processed),  10.0 o/o (all)
  Handled by content
        Discarded :    194 =  13.6 o/o (processed),  12.1 o/o (all)
     Quarantained :    128 =   9.0 o/o (processed),   8.0 o/o (all)
        Delivered :     71 =   4.9 o/o (processed),   4.4 o/o (all)


It also means that quite a lot of sites that used to send spam, found the site was not responsive and it seems they ‘forgot’ my domain. Which, of course, is fine.

relay attempts are minimized as well, the only ones I found, except the test ones that come from tiscali.it (from different IP addresses, by the way) there were a few in the 21st. Using my email address to a outlook user, and non-existing users to two different hotmail.com recipients.

For the rest: nothing weird.

Just the usual – with a little twist

Nothing wrong – again. Spam filter does it’s work, as usual:

PMAS statistics for May
Total messages    :   2783 = 100.0 o/o
DNS Blacklisted   :    106 =   3.8 o/o (Files: 25)
Relay attempts    :    177 =   6.3 o/o (Files: 31)
Accepted by PMAS  :   2500 =  89.8 o/o (Files: 31)
  Handled by explicit rule
         Rejected :   1400 =  56.0 o/o (processed),  50.3 o/o (all)
         Accepted :    225 =   9.0 o/o (processed),   8.0 o/o (all)
  Handled by content
        Discarded :    424 =  16.9 o/o (processed),  15.2 o/o (all)
     Quarantained :    327 =  13.0 o/o (processed),  11.7 o/o (all)
        Delivered :    124 =   4.9 o/o (processed),   4.4 o/o (all)

Twist is that 3 antirelay logfiles seem to be 800 blocks in size: 01-May, 05-May and 13-May, but these were deleted, where they should have been kept for examination. They could all be undeleted using DFU, but I found they were actually < 4 blocks in size, so it's clear why these were deleted. So deleted them, telling me each of them was 800 blocks - allocation size. Not sure what made them that big... Well, the system is a bit slow at times, perhaps this is symptom of something else I have to dig into. However, I will change to the second system some time soon, but I will have to take out the SA5300a card and put it into the other one (or just the cache card, that's broken on the other machine).
New hardware
On the workstation front: The new system worked fine but I ran into shortage of video memory of the RTX3080 card – thought 10Gb would be enough but it wasn’t. So I found RTX4090 card, 24Gb VRAM, got that and installed it instead. Now I may run into similar issues in the future but for now, this will suffice.
The old motherboard – on which the OS is installed on an M2 card – could be placed in the case I had this system originally, added a few hard drives that are now available (No room for them in the new case) and placed my good old GTX1080 in it. Booted from scratch with no issues. As soon as I get my community license for this instance, I will use this to run VMS on x86 on this machine, as a guest on VMWare Player (or Virtual box, not sure what, as for now) to replace the Alphas.

All quiet on the [western] front
Nothing new on the Alpha front. Even mail:

PMAS statistics for April
Total messages    :   2763 = 100.0 o/o
DNS Blacklisted   :      0 =    .0 o/o (Files:  0)
Relay attempts    :    127 =   4.5 o/o (Files: 30)
Accepted by PMAS  :   2636 =  95.4 o/o (Files: 30)
  Handled by explicit rule
         Rejected :   1606 =  60.9 o/o (processed),  58.1 o/o (all)
         Accepted :    152 =   5.7 o/o (processed),   5.5 o/o (all)
  Handled by content
        Discarded :    284 =  10.7 o/o (processed),  10.2 o/o (all)
     Quarantained :    281 =  10.6 o/o (processed),  10.1 o/o (all)
        Delivered :    313 =  11.8 o/o (processed),  11.3 o/o (all)
Stopping any PMAS detached processes....


Well, just that the antirelay log of 21-Apr-2023 was 800 blocks in size but apparently, EOF was less than 4 blocks. So it was deleted in the job. I guess it was still o-pen at that time.
Also, today I found that the BIND service wasn’t active. At least, not in $ SHOW SYSTEM but Outlook didn’t find the designated server in the network. $ @SYS$STARTUP:TCPIP$BIND_STARTUP noted the service was active. So starting with $ @SYS$STARTUP:TCPIP$BIND_SHUTDOWN cleared that condition, and after $ @SYS$STARTUP:TCPIP$BIND_STARTUP the server could indeed be found. The fact that the service didn’t response could be the reason that any internet access was slow on any system on the LAN…

“Quiet” doesn’t mean “no new”. Since VSI opened commu8nity access to VMS on X86-64, I requested a community license but since demand seems to be high, it will take some time. If that comes, I’ll need to set up a server to replace (in due time) the Alpha and Itanium systems. I’ll keep them for some time, though, just to have at least one system that works. Currently, it is Diana only, since the RAID controller of Dione is still non-functional: I found out that it is the cache module that broke down and a replacement is still not available. Contacting IslandCo via their gmail address is a problem – since VMS mail has no facility to connect securely (no TLS…) so I have to take another turn: Using either my gmail address, or my ISP mail facility. I still want to switch systems though, so I’ll have to do some more work: Read all mail from Diana, stop the mail front-end (so I won’t get new mail), dump the MySQL database and save it elsewhere, do all maintenance tasks (that is normally done in the maintenance job on the first of each month…), shutdown, move the cache module (of the whole RAID controller) to Dione and start that one; next update the database….
Quite an exercise… So hopefully, the X86-64 license will be received soon, so I can get that system up and running.

Also, on the hardware front, there is news.
As stated last month, the ASUS TUF Mark1 needed to be replaced for the majority of work: Since this at times requires a lot of CPU power, it was anticipated a more recent system would be an advantage. So I purchased an ASUS TUG Gaming Z790Plus motherboard, Intel 13900F processor and 64 Gb of DDR5 memory; installed 2 4Gb hard drives and 2 2Gb M2 boards (still 2 slots open); Used the original 1Gb SSD that was the Windows 10 system disk of my previous machine (even older) and left the 2Gb HDD that was installed in the case. This system was upgraded to Windows 11 and I installed all software I needed. This worked fine, until I removed the Windows.old directory, after which sound din’t work anymore – and there were issues with some other stuff. After re-installing Windows 11 from scratch on one of the M2 boards, and installing all software a second time, it now works as intended.
But I didn;t get rid of the TUF Mar1 board: It is still complete, so I installed it in the original case, including DVD and Blueray drives (that cannot be installed in the LianLi case) and two HDD;s that I had on shelf. This system is to be used as local file- and media server – and will host the VMS X86 installation.

Again – no surprises
Took some time, but checked the maintenance run of Fool’s Day – but nothing to mention:

PMAS statistics for March
Total messages    :   2521 = 100.0 o/o
DNS Blacklisted   :      0 =    .0 o/o (Files:  0)
Relay attempts    :    103 =   4.0 o/o (Files: 31)
Accepted by PMAS  :   2418 =  95.9 o/o (Files: 31)
  Handled by explicit rule
         Rejected :   1582 =  65.4 o/o (processed),  62.7 o/o (all)
         Accepted :    178 =   7.3 o/o (processed),   7.0 o/o (all)
  Handled by content
        Discarded :    285 =  11.7 o/o (processed),  11.3 o/o (all)
     Quarantained :    182 =   7.5 o/o (processed),   7.2 o/o (all)
        Delivered :    191 =   7.8 o/o (processed),   7.5 o/o (all)


Two-third of the messages rejected – meaning PMAS does a good job on the sender, domain, subject and content where I put a filter on to reject the messages (which means: If PMAS signals the message header of content matches a mask, the connection is aborted), and what was not rejected matched built-in rules (so the message is still passed but into “discarded” or “quarantained” folder, depending on the severity. The license needs to be renewed in May – no cost so I’ll do that anyway.

No much on relay attempts – just one (03-Mar-2023) that had a bit more data: 6 lines of test by Italian ISP – that one comes in almost every day) and this time a few others sending to hotmail.com. Of course, using “attempted” users in my domain that do not exist, and using an address that is not mine:

3-MAR-2023 10:24:12.31|R|185.246.220.193|remote@grootersnet.nl|zzzzzzzzzzzzzzzzzezihe@hotmail.com|550 5.7.1 Relaying not allowed: ♦
3-MAR-2023 10:24:12.52|R|185.246.220.193|remote@grootersnet.nl|zzzzzzzzzzzzzzzzzezihe@hotmail.com|550 5.7.1 Relaying not allowed: ♦
3-MAR-2023 17:18:17.53|R|185.246.220.193|user@grootersnet.nl|zzzzzzzzzzzzzzzzzezihe@hotmail.com|550 5.7.1 Relaying not allowed: zz♦
3-MAR-2023 17:18:17.79|R|185.246.220.193|user@grootersnet.nl|zzzzzzzzzzzzzzzzzezihe@hotmail.com|550 5.7.1 Relaying not allowed: zz♦
3-MAR-2023 19:26:12.44|R|193.47.61.35|driver@grootersnet.nl|workhouselog@hotmail.com|550 5.7.1 Relaying not allowed: workhouselog@♦
3-MAR-2023 19:26:12.68|R|193.47.61.35|driver@grootersnet.nl|workhouselog@hotmail.com|550 5.7.1 Relaying not allowed: workhouselog@♦


According WHOIS both addresses refer to a Dutch server provider, that will be notified – of course.

New hardware
Though the system will do it’s job, I found that ASUS TUF Mark1 combined with intel 9820 CPU (3.3 Ghz) and 64 Gb memory (capable of 3400 Mhz) has some trouble in compiling GEAE landscapes (which is CPU intensive) and Unreal 5 rendering. So I purchased a new motherboard: ASUS TUF Gaming Z790 + Wifi, Intel 13900 CPU and (again) 64 Gb memory, capable of 6200 Mhz. I could have gone for a 13900K – even faster but that is a lot more expensive and requires 100 Watt more power, so in that case I would need another power supply with higher output as well….The current machine will have it’s use when VMS -X86 is available for hobbyists – the Alpha’s and Itanium machines will be obsolete at that point. Furthermore, since my latest case does not allow DVD drives installed and the previous one does, it will function as a file and media server as well: it has enough power and cooling to handle that. Have to do some shuffleing with data, because there will be new disks as well with the new motherboard, the old drives are still in use (and need to be later on)