Linux

A few months ago, I was unable to login on my vCSA. At that time, I thought it ws a glitch, I rebooted (didn’t fix the issue) and changed the password of the users, which fixed the issue. Recently, I had the same issue. I concluded it could not be a glitch anymore and decided to search for root cause. The reason given by Veeam was “username or password incorrect”, which was wrong. On the other hand, vSphere Web Client gave another error, much more helpful : “account locked out”. ...

A simple script to check Exim queue size for Nagios NRPE: #!/bin/bash #arg1 = warning count #arg2 = critical count if \[ -z "$1" \] || \[ -z "$2" \] then echo "Usage: ${0} warningcount criticalcount" else if \[ "$1" -ge "$2" \] || \[ "$1" -le "0" \] || \[ "$2" -le "0" \] then echo "Error: incorrect values." else queuesize=$(/usr/sbin/exim4 -bpc) if \[ "$queuesize" -gt "$2" \] then echo "CRITICAL - exim queue size: $queuesize" exit 2 elif \[ "$queuesize" -gt "$1" \] then echo "WARNING - exim queue size: $queuesize" exit 1 else echo "OK - exim queuesize: $queuesize" exit 0 fi fi fi

On two recent Supermicro servers, typing reboot actually shuts down the server. Thanks to Google and this blogpost, http://major.io/2013/06/03/supermicro-x9scix9sca-server-does-a-shutdown-rather-than-a-reboot/ I no longer have this issue. Short story: just blacklist the module named “mei” and you’re good.

If you want to boot with DHCP/PXE/TFTP method and your DHCP is handled by a Mikrotik router and your TFTP server is on another server, then you need to set the next-server and boot-file-name parameters in the network definition. Do not try to set DHCP options (codes 66 and 67) and assign them to the network definition, it will not work. The same applies to the NTP servers option. So, do not do this: ...

This weekend, I finally had time to install a few network cameras from Ubiquiti in my house. The goal was to replace the previous system, which was composed of cheap USB webcams and the software motion , that has been running for years. HP Proliant N40L I had a spare N40L lying around that I used as a storage server for my VMware lab, so I repurposed it to be my future one and only home server running file sharing, monitoring, etc and AirVision of course. ...

It has been explained many times over the internet, so I won’t cover that. One thing you have to pay attention to is, if you run snmptt as a daemon, be sure that it has the right to write to Nagios FIFO. On my Debian system, I had to run snmptt as nagios user. Then, when you convert the MIBs to snmptt.conf file format, sometimes it will fail because the MIB is not RFC compliant. Most of the times, object definitions contain underscores (_) or dots (.) . Just sed them. ...

After a few hours spent trying to find why Ubuntu 12.04 would not download a preseed file while netbooting, I stumbled upon this bug https://bugs.launchpad.net/ubuntu/+source/casper/+bug/923219 . I needed to add “IPAPPEND 2” after the APPEND directive in the pxelinux config file. Great.

Last week, I saw an huge “spam” outbreak on my MX servers. Instead of ~200k rejected SMTP connections, I had ~3M connections and it was rising. This caused some minor issues with the frequent log parsing as the log file quickly grew. To reduce the log file size and also the DNS requests number, I decided to use fail2ban to drop the connections before they got to Exim. This is rather easy, just create/append /etc/fail2ban/jail.local with the following: ...

If you use Varnish with an Apache server and analyze Apache log files (with AWstats or something else), you probably saw all requests were coming from 127.0.0.1 (or the IP address Varnish is on). You’ve probably forgotten to change the LogFormat you use. Fortunately, Varnish adds an header on its request to Apache with the remote IP address in it. Just replace “%h” with “%{X-Forwarded-For}i” in the LogFormat and reload your server. ...

I needed to install some incoming mail servers (MX servers) to filter emails before delivering them to customers servers. So the need is to accept mails for a fixed list of domains, check for viruses and spams, then route them to customers servers for delivery. The distribution is Debian 6, the mailserver is Exim. The antivirus check is done by ClamAV and the antispam with Spamassassin + Pyzor + Razor + DCC + DNSBL. I get the list of domains from the f*cking OSS/BSS/whatever software. ...