server testing progress reports


Did well for two and a half months then two local power failures in two days both outlasted by UPS reserve and when the server came back up it wasn't allowing anonymous access to the web site. Don't know whether that was hacking or just really bad luck with the power failure. Sorry to those who were locked out.


Came back from leave (20th to the 27th) to find that the WWW server service had stopped and wouldn't be restarted. Then found that the user manager had decided I needed to change my password but also that I didn't have the permission to do this. Ah well, fooled NT into thinking it was May and it let me back in, still can't see why the service failed so thoroughly nor why the user manager got so confused -- which is worrying. Sorry for those who wasted time trying to connect during that period. Search for funding for a machine I can devote to the server goes on (hint, hint). Have added new conference information and updated some pointers to other things.


Server has been up continuously for about five days apart from a quick restart yesterday afternoon to assimilate new program mappings put in the registry (at last, the server can see my Perl script's output!).


Came in to find the server had stopped again last night after I left shortly before midnight. There seem to be three possible culprits: backup, my ptrs2.prl script running as a scheduled event, or my virus scan running as a scheduled event. I've dropped the two scheduled events from the scheduler and will run them now to see if they stop the server when run manually. I _HATE_ problems like this! (and am sorry for any inconvenience). Incidentally, anyone out there know how to get both anonymous and non-anonymous FTP working with the M$ IIS FTP server? Anyone know how to get perl standard output visible to the M$ HTTP server?


Computer unit changed the chassis on the hub I'm on yesterday (13th) lunchtime which knocked out half an hour or so as the port was defective when everything was restored. Worryingly, IISA monitoring didn't appear to notice anything wrong at all. Then that threw an exception error at midnight which I don't understand at all. That meant the server went down (or ) until 08.15ish this a.m. when I came in and found everything static. Just when I was starting to trust the thing! Otherwise no major problems and a lot of steady development work on the site.


Finally worked out how to get the actively updated statistics on usage focussed on the right days (!) and added new information on research utilities, q'aires, measures directories(developing)


Server down briefly yesterday afternoon when I was sorting out a problem on the serial port. Looks like it will have to go down for replacement of two fans and to install the NT service pack 4 to get the MS dbserver up for trial. I'll try to make sure I give notice of downtime in future. Also down momentarily for installation of the MediaHouse IIS Assistant which is now providing me with a monitor of who's on the system which should make the fall out from taking it down less serious in future. That is also providing actively updated statistics on usage and will be one of the sources of file usage statistics when I can find time to sort that out. Until then, the pointers to statistics will continue to point nowhere. Sorry!


Got two search engines up and running, but neither configured exactly as I'd like: the Excite engine and the Verity engine which comes with three search interfaces:

17.vii.96 16.15

Server up and running as of about now. Sorry for the down time (for those who hit it). Sometimes NT can be less than entirely friendly and adding a second processor gets into the guts a bit!

18.35 Tuesday 16.vii.96

New drive and extra RAM installed. Will have to delay adding second processor as MS, in their wisdom, now make the utility to convert the NT 3.51 server from uniprocessor to multiprocessor configuration an extra cost option within the resource kit and no-one told me. So ... one more downtime scheduled for later this week (I hope). ...> well, that took me down for most of today (17.vii.96) ugh!


...This is the new service from my NT machine. Today has been the 14th of July 1996 and it's not been a good one technologically! Having put the server up and running in what is really an alpha rather than a beta test I had to take it down to reboot to see if that would sort out some problems. I decided to do both low level (SCSI) and high level (chkdsk) checks of all the drives to make sure they're O.K. It seems they are! I also noticed that all the links in the system were wrong which must have been pretty galling for anyone who checked in for me. I'm sorry if that was you!

That led me to a currently unfulfilled search for a proper unix find function for NT (the posix resource kit version from MS doesn't seem to work, nor does a GNU version I downloaded from a German university server. If you've got a general set of "Unix for NT" whose source (preferably i86 binaries zipped) you can give me I'd love to know.

However, on with the testing. Please let me know about any broken links that appear or any other problems. I'd also like to know from experienced users how the response rate compares with that of
All information to: please!
WebTechs Mozilla Checked O.K.! 5.viii.96
KGV Netscape-HTML Checked O.K.! 5.viii.96