1. vmail: a) HP Webmail - first 10 points are performed and done. - 2nd round of queries are fulfilled also. 09-09-02 b) horde workshop - initial stageof getting horde from cvs and automatically changing the .dist file to originals is done. - 2. new blackbox: a) SENDMAIL: - creating of the sendmail.cf with all the necessary macro definitions suited for blackbox. - compiliing. - initial setup is done. - install makemap, - it needs to be configured i.e. the initial scripts, access database and other database. .. in /etc/mail access, virtusertable, mailertable, local-host-names, domaintable, aliases etc. files .. in /etc/init.d sendmail startup script .. in /etc/rc2.d /etc/rc3.d start script with priority settings .. - test, test, test, test, test .... b) UW IMAP & POP (Remove QPOPPER from the machine) c) XTRADIUS d) POP AUTHER e) Script to add users (2000+) to the system (not copy). f) Script to create the /home/ directory. g) 3. changing scripts in Netman, Flow and Corporate to ssh instead of telnet. DONE on 25-09-02 4. write a script to add spam emails in the access database and makemap the database. - in main:/maintenance/bustspam.pl - almost done, basic modules are done. - pending work is: reading from file, reading from list and doing the actual thing. 08-09-02 8:45 pm - idea: why not let the program read from a default file and add the spam list from it to the access database everytime the program is invoked. 08-09-02 10:00 pm - ok the idea is implemented, but now there is a iteration problem. which is actually comparing two file. My brain is not working or else this would be solve in no minutes. 10-09-02 5:50 PM - Added the code to ignore the remarks in the /etc/mail/access file. And, fixed the iteration problem. 12-09-02 1:26 PM - Added the duplicate entry check for -a option, but its not working properly. Probably, there's a problem in the pattern matching module. 13-09-02 5:40 PM - Ok fixed the problem with the -a option and duplicate address containing '*'. Never worked on this till today after 13th :). 20-09-02 1035 AM # #5. write a perl script to detect cients ip address for the gnudip client application that willis is writing. # # - /usr/local/apache/cgi-bin/ip.cgi # DONE # - almost done, can detect the IP that is connecting to GNUDip # # - Can't detect the IP address that the browser is running from. 09-09-02 5:59 PM # # - Changed the code to give a neat and clean HTML. And willis only wants the computer ip. So this project ends here. 19-09-02 6. Install SOLAR2 from scratch, as its hardisk crashed. 26-09-02 - Let Alex know about the harddisk crash. 26-09-02 -DONE- - Install new harddrive. 26-09-02 -DONE- - Save the data in the other drive if possible. 26-09-02 -WAS-NOT-POSSIBLE- - The data in the other drive could not be saved because the new OS does not recognise the partition. 26-09-02 4:45 pm - New OS is installation. 26-09-02 5:00 pm -DONE- - Configure the OS for basic operations. 27-09-02 9:30 am -DONE- - Repartition the 2nd hard disk as according to Solar1. 27-09-02 10:00 am -DONE- - Make the system look alike of Solar1. 27-09-02 10:30 am .. /etc/rc$ dirs should have the same service that solar1 has >> /etc/rc2.d/S88httpd, /etc/rc2.d/S88netsaint, /etc/rc2.d/S88sshd -DONE- >> /etc/rc3.d/S80mysql, /etc/rc3.d/S86radiusd, /etc/rc3.d/K80mysql -> ./S80mysql -DONE- .. check whether any system specific configuration has been made in solar1 or not. If yes, do the same in solar2. .. chack the partitions are same or not. .. check what are the service and directory conventions used in solar1 and jot them down. >> httpd - /usr/local/apache, version: Apache/1.3.14 (Unix) >> radiusd - /usr/local/radius, version: Radiator 2.17.1 >> mysqld - /usr/local/mysql, version: mysql Ver 11.12 Distrib 3.23.33, for sun-solaris2.8 (sparc) >> netsaint - /usr/local/sbin - Now check the archive for solar2 in wave and match the services that were installed from the /home/admin1 with solar1. .. we only have mysql's var/ directory. so, gotta copy the mysql from solar1. -DONE- .. we have the whole radius directory. so restoring the whole dir with do. -DONE- .. we have the /home dir, so we can just unpack it. -DONE- - There is a second IP for solar2, assigned. NEEDS A REBOOT. 30-09-02 4:45PM - After copying the mysql and starting it, it seems that mysql is running fine. Even the databses are verified nicely. Now the problem is the Replication settings. *** THE REPLICATION SETTINGS NEED TO BE VERIFIED *** 30-09-02 5:45 PM .. check the replication setting are there or not by SHOW SLAVE STATUS in mysql client. >> this command returned that slave is running. GOOD sign. BUT HAVE TO RECHECK WHETHER THE REPLICATION IS REALLY HAPPENING OR NOT. .. check the var/my.cnf for the master entries. >> fine and ok. - After copying the radius directory and i ran a test in the radius. ALAMAK! It does not verify the username and password from the database. *** NEEDS AN EXTENSIVE TEST AFTER A REBOOT *** 30-09-02 6:15 PM - Examine what are the other scripts or services running in solar1 and start them in solar2. 01-10-02 11:46 AM .. /usr/local/sbin/netsaint_statd -DONE- - Examine the crontab in solar1 and make solar2's crontab same, but DO NOT ACTIVATE THEM. 01-10-02 11:50 AM .. /usr/local/bin/logcheck.sh -DONE- .. /usr/local/scripts/dumpmysql.sh -DONE- .. /usr/local/bin/glogs.sh -DONE- - REBOOT THE SERVER BUT FROM THE DATA CENTER AS THE BOOT TIME MSG HAS TO BE MONITORED FOR ANOMALIES. 01-10-02 12:30 PM - OK. After REBOOT it seems everything is working fine. But there's a repeated error message in the radius log of the 'radius' user cannot connect. And that is happening for 'qalatest' user. I digged into this and found that this started from the June '02. -STRANGE- Other than that everything seems to be just fine. 01-10-02 4:00 PM - Check the mysql replication process working or not. Key commands: 'show slave status' (if Slave_Running is Yes) 'show processlist' (if the State is anything but 'connecting...') then the replication is fine. 01-10-02 4:10 PM - If everything is well and running, add the ssh-key for wave to solar2 so that wave can login and backup solar2's data again. -DONE- 01-10-02 4:18 PM - LAST TEST: Add a user in soalr1 in SUBSCRIBERS table and check whether the user is replicated in solar2 or not!!! --------------------------- TEST SUCCESSFULL ---------------------------------------------------------- 01-10-02 4:32 PM - Send a mail to Alex and everyone who are concerned that SOLAR2 is up again. 01-10-02 4:52 pm 7. Blackbox QUOTA Support. - As Blackbox has only 9.2 GB of hard disk space, space has become a very big problem. To temporarily prevent this I will install QUATA support in Blackbox. - As QUOTA support needs a kernel update and a kernel parameter to be changed, its a bit risky to do that. As for a Production system I do not want to take that risk, I will compile the kernel from scratch and install a different instance of it and load this kernel with QUOTA support in one of the Maintenance Night (may be 5th Oct '02) and check whether everything is working file or not. - If everything is working fine, then the kernel will have to be the default kernel. - Now the checklist: ... Backup the current kernel. -DONE- 01-10-02 5:18 PM ... Make changes to the LILO configuration to have a backup booting kernel. -DONE- 01-10-02 519 PM ... Change Kernel parameter for QUOTA support. ... Compile Kernel. ... Check whether the current kernel is replaced with the new one or not. ... Change LILO configurations again to load from the Backup kernel by default for pre-causetion. ... Backup the /etc/fstab file. ... Change the /etc/fstab file to give the QUOTA support in the desired partitions. ... Create the two empty files /aquota.group and /aquota.user ... Wait for the maintenace day ... ZZZzzzzzz... :) ... On Maintenance Day: >> Change the Lilo configuration to make the with QUOTA kernel the default kernel. >> REBOOT >> Watch for any errors. >> If no error, check the size of the /aquota.group and /aquota.user files size. If they are Zero then there's a problem. >> Enter 'quota' to see whther it gives any error or not, if not we are ready to set the quotas for all users. >> Create a default quota for user root with edquota. >> Now copy the root's quota configuration to 2000+ users quota configuration. Zoooom... >> Voila done. 8. SOLAR3 installation. Make a MIRROR of SOLAR1. - Insall Solaris 8 in the 220R (typhoon). (Benchmark the exact time of installation for records) - Do normal configuration and hardening. ... Get rid of unnecessary services. ... Close all ports except the ones needed. ... Make sure all security patches are installed and any official patch from Sun. - Install services and packages: ... Install essential packages: >> GCC - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/gcc-3.2-sol8-sparc-local.gz >> Perl - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/perl-5.8.0-sol8-sparc-local.gz >> TCP Wrapper - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/tcp_wrappers-7.6-sol8-sparc-local.gz >> Wget - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/wget-1.8.2-sol8-sparc-local.gz ... Install SSH Prerquisites and SSH: >> EGD - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/egd-0.8-sol8-sparc-local.gz >> PRNGD - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/prngd-0.9.25-sol8-sparc-local.gz >> ZLIB - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/zlib-1.1.4-sol8-sparc-local.gz >> OPENSSL - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/openssl-0.9.6g-sol8-sparc-local.gz >> OPENSSH - ftp://ftp.sunfreeware.com/pub/freeware/sparc/8/openssh-3.4p1-sol8-sparc-local.gz ... Install the Security Package that is recommended by SUN. ... Install the required services: >> RADIATOR >> MYSQL >> GRIC >> iPass Netserver