Hey JP folks,
Sorry to follow up via email chain but for me its easier for me to
organize my mail via threads given I have about 3 dozen alias emails
going into different boxen via gmail aliases.
I just wanted to comment on the most recent LUP regarding Linux malware
and protection. I run hundreds of servers, many with well know public
IP's (ips from decades old running class A networks...yeah those class
A's....). I have some straight forward suggestions that have kept my pub
facing bits from getting pwned.
I would be happy to elaborate but basically:
1. Fail2Ban
2. sshd_confid "permitroot without-password" - only allow root ssh
logins when using private keys
3. Ossec/wazuh or other log reading aggregators for when you miss things
I have been told that many people turn off selinux on their servers
during their kickstart installation script....I would not know about
that though........................but it hasn't hurt _them_ yet....
Also, would love to know what you folks think about Ansible vs Salt. I
have used both but am more of a salt fan myself as it allows our
contracting company to reuse states between clients due to salts pillar
system.
Thanks for all you do!
Morgan
P.S. The server I sent you guys is one of tens of thousands that run
with inlet temps around 80 deg F so as long as you keep them under 90
inlet temps you should be fine, even if they occasionally throttle.
P.P.S. I work in a datacenter and am incredibly close with the facility
director who designed it if you need any advice on cooling/making a cool
rack. FWIW, I would suggest a box with at least a big ol' hepa filter in
the front. At the end of the day, HEPA filters are crucial in our
datacenter.