Month: March 2008

WebLogScan update
There were some slight issues with the scanner of webserver access logs: some valid locations showed up to be invalid…So I changed the search sequence in the routine and enhanced it in some other parts. The log now contains invalid locations only.
Wiki
Digging my way through the MoinMoin configuration options. There is still a lot to do getting things right: authentication, setup a second (public) wiki aside the test one (which is still in the restricted area). The documentation in the official site is based manliy on Unix and Windows – not really a surprise – but when the pages are available offline to be examined, it would make things easier. Perhaps JFP has some more on his site. We’ll get it online, no doubt. I would like to use it for documenation.

and for the rest – another update has come along to be downloaded – and installed.

Wiki running
After some discussion with Jean-Francois Pieronne, a new container file is downloaded that holds adjusted files so I could be able to run a wiki (moinmoin, in this case). He also made adjsutments to the online documentation and gave some hints. Tonight I took some time to see if that would be the solution.
One big change, though: It now uses Mark’s PyRTE – the Python RunTimeEnvironment.
Well, partly.
Since I run it in the oprerator site to begin with, the mapping for the webserver must be chnaged slightly: it contains a line that allows the RTE to run under a different user – and JFP nicely used http$nobody this time, but still this function is disabled (due to HTTPS…). No problem,though.
The first attempt failed, but that was because I forgot to change protection of the command procedures to W:RE. After having done that, it took a long time – but the frontpage did indeed show up. VERY basic, however, no formatting, images and so on – it looked as if stylesheets were missing. Next, I changed protection of all files to W:RE and retried. Now the pages showed up properly. I created a user (myself), next I’ll have to do some editing and checking.

The documentation howver still fails, and I notified JFP. It looks as if a file is missing – it does exist but in another directory than expected. This is still to be handled.

WordPress 2.3.3
The Blog on the 2008 bootcamp is set up using WordPress 2.3.3 – and this seems to require a bigger amount of memory, compared to 2.2.3 that this blog is using. I added an entry, and when acecssing it to edit is, MySQL crashed – again. Once restarted – after some time – editing the pages was faster then before and MySQL kept up. So the addition could be made. It seems the longer MYSQL is up, the slower it gets and the more likelier it is to crash.
I wonder if 5 would be better.

PHP trouble
This entry could not be published when I wanted to: the PHP engine crashed:

%HTTPD-W-NOTICED, 13-MAR-2008 23:18:26, CGI:1969, not a strict CGI response
-NOTICED-I-SERVICE, http://www.grootersnet.nl:80
-NOTICED-I-CLIENT, 192.168.0.33
-NOTICED-I-URI, POST (26 bytes) /sysblog/wp-admin/post.php
-NOTICED-I-SCRIPT, /sysblog/wp-admin/post.php sysblog:[wp-admin]post.php (cgi_exe:phpwasd.exe) SYSBLOG:[wp-admin]post.php
-NOTICED-I-CGI, 2553595354454D2D462D485041524954482C206869676820 (129 bytes) %SYSTEM-F-HPARITH, high performance arithmetic trap, Imask=00000000, Fmask=00000002, summary=02, PC=00000000001E9C94, PS=0000001B
-NOTICED-I-RXTX, err:0/0 raw:3586/0 net:966/0

time after time. So I retried half a day later.

Not enough core
Yesterday evening, the WordPress code signalled this error on the dashboard – the admin entry page – and subsequently, MySQL crashed as well, probably with the same error. At the time (23:00), I decided to add the entry next morning and look into it later.
So this moring, I logged in as admin – and again, “not enough core” in PHP and again, MySQL crashed – I cannot tell whay at the moment – I’m not in the center – but quite likely it’s the same issue as yesterday. Luckily, MySQL is revived automatically by it’s watchdog. A second attempt gave the same error in PHP but this time it looks like MySQL survives this time.
Watching the system, there is no appearant reason for memory consumption. I’ve been busy with MoinMoin, but only a short period since external system management had a higher priority: a friend’s PC had trouble when connecting to the network, it looked to enter standby mode when a network cable was connected. The cause was twofold: the video card lacked power, and the network card’s drivers were missing. This is now taken care of, and although there is still some work to do to make the system fly again.

MoinMoin
Jean-François Pieronne (who did the port of this wiki-code) read about the problems and
gave me some hints. One thing I want is to follow the same method that I use for the blogs – and that requires the definition of a number of logicals. I did define WIKI now, and assured myself that the files required were available: One thing that might have been wrong is exactly this file. A moincgiplus.template is delivered – and I copied it to moincgiplus.com and looked where I had to change things. I couldn’t find anything and so I just tried. To no avail: there still is this problem…
— To be Continued —

But this should not have caused the memory exhaustion…

Updates ahead
The latest consolidated VMS update has been downloaded, it contains all updates since the last one, and none is excluded. So one of these days, I’ll do an update of Diana.

No Internet connection
yesterday, I scanned a few documents and when done, I left the working plase, after having the scanner disconnected from the power grid.
At least, that was what I thought.
Only this morning I found out – by pure accident – that there was no Internet connectivity. No wonder: I had taken off the wrong power supply: the one of the router.

So there has been no connection between yesterday 19:30 and this moring about 6:30.

Time to re-locate this connection to a place less easily reachable! This happens too easily.

No problems on the sytems. A quick scan on the processlist using the server’s logging facility showed everything is still up and running, though in first grance it looked as if the spamfilter was gone; but it was just that I overlooked the processes. All processes are running, but some are swapped – likely because I logged in uisng the XWindows interface on the direct attached monitor.
The system’s performance graph missed out all spikes that occur every hour or so – probably a webcrawler.

The only thing is that mail has not been delivered, and though my ISP states they hold a backup (Store & Forward) chances are it didn’t work. I’ve seen that before.