Bartosz Feński: step closer to upgrade of my server
I've got a new second hand hardware for my server, so I'm step closer to upgrade it eventually. Right now server is running on an old Celeron 300MHz, 256MB of RAM and 80GB hard disk.
(fenio@serwer)~$uptimeAs you can see it's very loaded. Mainly because of Apache which eats almost whole CPU power serving about 40 domains. In addition almost each domain have virtual mails, ftp and sql base. So it's about time to upgrade it. New hardware is 2xPentium III 733MHz, 1GB of RAM and 2x250GB disks. I hope that'll be enough for some time. I didn't decided yet how to use disks... LVM comes to my mind, but maybe RAID since I've got hardware controller for that? RAID0 seems to be interesting solution and should increase overall performance. Changing hardware seems to be great opportunity to finally switch from Woody to something newer. Am I last person still using Woody these days? :) I'm going to finally use some software that'll allow me to put some duties on users instead of me. I mean some software helping administration of virtual hosts. I found three interesting projects for that purpose: Any experience with some of them? Any recommendations? Or maybe something else I missed during search? My personal wishlist is:
14:19:07 up 59 days, 21:57, 1 user, load average: 10.70, 10.75, 10.61
(fenio@serwer)~$
- DFSG compliant
- co-op with bind (I can use other DNS daemon eventually)
- co-op with postfix
- co-op with proftpd (pure-ftpd eventually)
- co-op with some webmail solution... openwebmail? squirrelmail?
- co-op with mysql (postgresql eventually)
- co-op with some pop3/imap servers... anything that's able to work with virtual boxes
- handling of different limits (number of subdomains, mails, aliases, accounts, etc.)
- complex statistics of everything
- easy installation on Debian
Edit: the same statistics on our backup/secondary MX
The green area are the mails we accept for delivery (after the AV-checks and classicale anti-UCE-checks). The red line is the amount of messages that the bayesian antispam (bogofilter) marks as spam. Since our clever greylisting is up, the results are just unbelievable: we drop a huge amount of spam, without needing bayesian spamfilter or clamav or any CPU-intensive checks, so almost for free.
The statistics we run show that a very few mails are actually submitted again (less than 20 per week), and most of them are spam anyway. And an analysis of the remaining spams shows that those are spam that come from redirections services (mail lists, @debian.org, ). We almost drop 99% of the spam that is submitted into the SMTP network through one of our MX's.
For the record, we only use cbl.abuseat.org, dul.dnsbl.sorbs.net and dynablock.njabl.org.
why doing it that way ?
greylisting is all about creating a dynamic map of the SMTP queues of the planet.
What I do, is only using RBL's that already exists to make that principle converge faster. Moreover, even if you use "bad" blacklists that have too many false positives, the greylist is here to let legitimates mails through.
So basicaly, it combines two bad[
Sprited and boisterous!