06-Nov-2019

Some issues all around
Just not on monthly maintenance job, that ran fine. Just the mail is troublesome:
PMAS statistics for October
Total messages    :  98888 = 100.0 o/o
DNS Blacklisted   :    108 =    .1 o/o (Files:  1)
Relay attempts    :    605 =    .6 o/o (Files: 30)
Accepted by PMAS  :  98175 =  99.2 o/o (Files: 31)
  Handled by explicit rule
         Rejected :  81852 =  83.3 o/o (processed),  82.7 o/o (all)
         Accepted :    149 =    .1 o/o (processed),    .1 o/o (all)
  Handled by content
        Discarded :    259 =    .2 o/o (processed),    .2 o/o (all)
     Quarantained :    111 =    .1 o/o (processed),    .1 o/o (all)
        Delivered :  15804 =  16.0 o/o (processed),  15.9 o/o (all)

Again, two days of relay attempts – the same signature as before but using another server on both occasions (sender (random)@grootersnet.nl, intended recipient danivela1029@gmail.com):

 4-OCT-2019 08:11:07.13 -  4-OCT-2019 08:16:11.10 (289) 185.208.211.225
17-OCT-2019 00:10:18.20 - 17-OCT-2019 00:16:04.58 (290) 185.208.211.244 

This address range is owned by a hosting company in the US:
185.208.211.0 – 185.208.211.255 = Dedispot Web Solution private limited
that has been notified, also about the previous attempts via other hosting companies.

The huge amount of messages (just under 100K) where most were rejected, either because they are blacklisted (almost 16000) or by rule on SPF failure, or any combination I set up (the rest) shows that blocking of networks in the router doesn’t always work, so the spam filter is constantly battered with these messages. However, this time it had an unexpected side effect.

DNS issues
All of a sudden, there was no way to access any website from anywhere in the LAN – though using mobile network still succeeded. On no system, translation of site name to address was possible: no DNS servers could be contacted. Even on the router, pinging the DNS resolvers of my ISP failed, though the DNS servers could be accessed. At least, using IP4. Using IP6 to contact the DNS resolvers did work.
There has been no change locally…(no, really). Question: could it be a router issue? I don’t use the ISP-applied one, so to find out I tried to install that one, but I didn’t remember the access password and the one on the sticker didn’t work either; solution is to reset the router, but no data was found on the internet (via mobile phone), contacted ISP helpdesk that helped me out with that – but reset failed. Probably because this was a very old one, so they sent me a new one. With that, I retried after connecting a laptop and tried to access the resolvers: This time, that succeeded. Next tried the Draytek – which succeeded as well.
The mail statistics for these two days gave a hint on the period the DNS resolvers were not accessible: On 31-Oct after 02:00 there is drop in messages that is persistent – and after 09:00 the number of messsages that were actually handled was minimal:

That continued until November 1st, 16:00, when the first messages were accepted again, an after 23:00 it seems the problem was solved:

Without intervention… My guess is that at XS4All something messed up…
But since the router is also outdated, I ordered it’s successor. installation should be simple since I can use a backup of this one.
The aftermath
Part of this has been that I removed te resolver addresses from the VMS BIND configuration, using just the router (that should take care of all unresolved references). I had to restart TCPIP to get that all in place. But that didn’t work as planned to I rebooted the VMS box altogether. That turned out to cause a problem accessing the blogs – because the PHP environment wasn’t setup properly: It started PHP 7.2 – known to be still a bit problematic. So I got the whole thing setup properly only yesterday evening….

Windows 10 updates fail
My laptops are older as well: Still have an XP one and two that run Windows7 from the beginning, both dating back to 2010 or so. One of these has been updated to Windows 10 in 2016 and some updates could be installed, but getting from version 1511 to 1803 stalls at 75%, and fails on 81% – where the previous version is restored. I tried that several times, and yesterday I had a try to update to 1903 – that stalledas well as about 75%, and reverted after 90%. I’m now so fed up that I ordered a brand-new laptop, the one still on Windows7 ill be used as console (now done by the XP one) because it has a serial port (hurray – no more hassling with COM ports) and latest teamviewer can run on that one – it cannot be installed on XP.

PHP 5.2.13 –> MariaDB???
Problems usi9ng PHP 5.2.13 and MariaBD 5,5 keep causing issues, connection is constantly lost….