A News Server provides services of Usenet news, which is an important application on Internet. It consists of discussion groups, which talk about everything. Actually it is a place where you can participate in discussions and express yourself by sending articles to the newsgroups.
Usenet is built hierarchically. The upper layer contains the main-groups, like for instance the maingroup comp. Comp consists again of sub-groups, as comp.apps and comp.ibm. Also these groups can be divided in other groups.
Some newsgroups are `moderated'; in these newsgroups the articles are first sent to a moderator for approval before the article appears in the newsgroups.
With Netscape you can join newsgroups. New users of News should first briefly read the group, news.announce.newusers, which contains introductory information. For users who want to quickly know something about an existing subject on Usenet, it is best to read the group, news.answers. It contains the Frequently Asked Questions (FAQ) and the periodic, informative articles, which are archived at the news.answers archive.

For a few years CS department has had an own news server. The current software version on the server is INN 1.4 sec. The machine itself is a dated SUN 4/470 with 3 technical outdated disks of 1 Gb for news.

Benefits of having an own News Server are amongst other things:

• Establishing of own storage times of the news, because of this interesting hierarchies can be stored longer.
• To be able to have discussions, which don't go outside the CS Department.
• To be able to anticipate on new interesting groups and hierarchies.
• To be able to have limited available inter-local discussion groups in case of collective projects.

Hierarchy and defaults	number of days
Moderated default	14
Unmoderated default	7
utwente.*	nooit
inf.*	nooit
ucam.*	365
comp.*	31
ieee.*	31
gnu.*	31
nl.*	14
nlnet.*	14
hacktic.*	14
xs4all.*	14
dds.*	14
vpro.*	14
alt.*	3

However at the moment there are some temporarily adaptations introduced by diskspace problems:

Hierarchy and defaults	number of days
alt.*	1

Below the results of each `down' server are described separately. These descriptions are given in aphabetical order on the aliasnames of the servers.

Athena

Mail from and to the working group IS is not sent anymore. Systems, which are connected to the server, go down. Systems that uses Athena as NISserver, will then use Zeus as NISserver. Also the home-directories of IS are not available.

Atlas

This is the TIOS compute-server. Nobody is really dependent on this server, however longlasting and heavy test-jobs are done on this machine. Of course, this is not possible when it is down.

Cerberus

This is the Trese (Seti) X-terminal server and WWW-server. When the machine is down X-Terminal users cannot work at Trese and the service for WWW is also not available.

Cloacina

This is the NIS cannot be done. Systems connected to the server, concerning filesystems also go down. Systems, which use Gaia as NISserver switch to the server Zeus. Updates of password cannot take place, because Gaia is NIS-Master-server.
The SLCs from the general cluster completely depend Gaia, so these perform neither.
Also home directories of B&O, misc_lab, IAPC and InterActief are not within reach.

Hades

This is the backup server. The PC-backup file systems are not attainable and no restores can be done.

Hercules

The machines on Buronet, can not work anymore. Hercules is very important for the Buronet, because this machine is NIS-Master, print-server, file-server, etc. for BUROnet.

Hermes

Systems connected to the server go down. Systems, which use Hermes as NISserver switch to Myrtilos or Zeus. Hermes contains home directories of the following groups: aps, misc, secr, and stud. So these home directories are not within reach.

Homerus

This is the printer-server, so when this server is down, nobody is able to print.

Hydra

This is the X-Terminal server for a great deal of Seti. The X-Terminal users in Seti (excluding Trese) cannot work. Besides this machine is WWW-server of Seti (also excluding Trese). Then this service is obviously not attainable.

Myrtilos

Sending mail from and to TIOS is not possible. Systems, which are connected to the server go down. Systems, which have Myrtilos as NIS server switch to Hermes or Zeus. Myrtilos contains home directories of the following Unix groups: arch, cps, ctit, fm, kam, multi, tools and tst. These home directories are not within reach, when Myrtilos is down. Furthermore, the Myrtilos also contains local Unix applications for TIOS. The programs are not accessable, so this means that whole of TIOS cannot work.

Nemesis - Hera

This is the UNIX practical training (EDUnet) server. For the practical training server this is the UNIX fileserver, PC-NFS file- and loginserver, mail- and NISserver. When this server is down, only the Novell practical trainings within EDUnet environment work.

Poseidon - Perseus

Mail from and to SPA and Seti cannot be send. Systems, which are connected to the server go down. Systems, which use this server as NIS server switch to Zeus and home directories of SPA ans SETI are not available.

Utrhcs

Mail connections with sites outside the Computer Sience department are not possible anymore. Ftp.cs.utwente.nl is not within reach. Hostname resolving on base of DNS doesn't work either. Besides the server is our NTPserver, so the few people, who make use of this directly also get problems.

Zeus

All central installed software is stored on this machine. When the Zeus goes down, the software is then of course not available. The software concerned is stored under: /Window,/Text, /Language, /Dbase and /Utils and /Gnu for the Solaris 2.x systems. Machines, which use the server as NISserver , switch to their working group servers. This machine is also the PC-NFS login and boot server. This means that PC-NFS users are not able to login or boot. Also the printer structure, which is kept on Zeus, is for a great part dependent on this software. This holds for serial locked printers. Probably, printing of postscript documents on QMS printers is possible.

---