Open Source

Getting ‘FreeBSD-10.2 is vulnerable’ messages on a 12.0 host

I started playing with /usr/local/etc/periodic/security/405.pkg-base-audit as part of a monitoring system. It works fine from the command line, but when I use Nagios plugins, I am getting unexpected results. By unexpected, I mean messages about FreeBSD 10.2. The host in question runs FreeBSD 12.0. The problem cannot be reproduced on the host, only from the Nagios monitoring host. Oh wait, the Nagios monitoring host is a jail on the host in question. That […]

Getting ‘FreeBSD-10.2 is vulnerable’ messages on a 12.0 host Read More »

Double timestamps in logs

I noticed some double timestamps in my logs recently. They started just after I upgraded the host to FreeBSD 12, but I am not convinced they are related. This is from /var/log/messsages: Jan 22 21:41:40 knew 1 2019-01-22T21:41:40.760533+00:00 knew.int.unixathome.org pkg 89351 – – py36-iocage-devel upgraded: 1.0.0.20181219,1 -> 1.0.0.20190122,1 They started late yesterday, this is from /var/log/maillog: Jan 21 22:28:58 knew 1 2019-01-21T22:28:58.677083+00:00 knew.int.unixathome.org postfix/anvil 42521 – – statistics: max connection rate 1/60s for

Double timestamps in logs Read More »

Upgrading to FreeBSD 12.0 from FreeBSD 11.2 using beadm and freebsd-update

Today I will upgrade knew from FreeBSD 11.2 to FreeBSD 12.0. It so happens that this is my last server at home which is still running 11.2, but I do have another server still on 11.2, but that one is at NYI. This post isn’t so much about beadm or about freebsd-update. I have written about moving to a beadm layout, but if your zfs list output looks something like this, you’re good

Upgrading to FreeBSD 12.0 from FreeBSD 11.2 using beadm and freebsd-update Read More »

Collecting statistics from bind / named

I use bind (Berkeley Internet Name Domain) as my DNS server. I am currently running bind 9.11.5P1 on FreeBSD 11.2-RELEASE-p8 in a jail, with iocage as my jail manager. The OS, jail, and jail manager should play no part in how this works. I have been collecting statistics from bind for some time. I have configured LibreNMS to collect the details via snmpd and they are plotted in a lovely looking graph. The

Collecting statistics from bind / named Read More »

Configuring LibreNMS Nginx statistics

This post might help me in configuring net-snmp to deliver Nginx statistics to LibreNMS. As such, this post may not be as helpful to you as the official documentation. Key is this URL: http://127.0.0.1/nginx-status The nginx.conf file needs something like this: NOTE: the fastcgi_pass line might also be fastcgi_pass 10.80.0.92:9000 depending on what you find in /usr/local/etc/php-fpm.conf. You should see something like this in the script, which you might have copied from https://github.com/librenms/librenms-agent/blob/master/snmp/nginx

Configuring LibreNMS Nginx statistics Read More »

Adding a zroot pool to an existing system

Current FreeBSD versions will allow you to create a zroot zpool from which you can boot. However, when I created the main zpool for slocum (on Fri May 3 2:16 2013), that option did not exist. You might ask: how do I know that date? [dan@slocum:~] $ zfs get creation system NAME PROPERTY VALUE SOURCE system creation Fri May 3 2:16 2013 – At present, the system boots from a raidz2 zpool. This

Adding a zroot pool to an existing system Read More »

x8dtu

This is x8dtu (named after the SuperMicro motherboard). This is the new FreshPorts server. The older post is still available In short: FreeBSD 11.2 booting off a mirrored pair of zfsroot SSDs 4.5TB of mirrored ZFS 196612 MB of RAM (yeah, that’s 196GB of RAM) SuperMicro X8DTU motherboard Intel Xeon E5620 @ 2.40GHz (two of those, giving 16 CPUs) NOTE: this post replaces

x8dtu Read More »

knew

This post has been replaced by a newer post. This is the previous post for this system configuration. For future reference, this is the knew server … oh wait, I think it’s this server which is was mounted in the 3U chassis mentioned in this post (perviously erroneously referred to as a 4U). It runs a few jails, including Bacula regression testing services. It is now mounted in a SuperChassis 846E16-R1200B This recent

knew Read More »

zpool: FAULTED – too many errors

This server, knew has had an intermittent problem related to CAM status: SCSI Status Error messages. There is a FreeBSD Forums post about it. On Sunday, the problem returned, and this time it degraded the zpool. I collected the information in this gist and I will list the relevant portions below. I had enabled smartd testing and I received this email late on Sunday: Despite the promise of more information in the logs,

zpool: FAULTED – too many errors Read More »

degraded zpool

On Monday night at about 5:30 PM, I noticed the email: ** PROBLEM alert – knew/zpool is CRITICAL ** What sickened me was that the email was sent at 3:58 PM. That makes me think I should add a Pushover.net alert …. I could do that at my mail server. Logging into the server, I saw: [dan@knew:~] $ zpool list NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT system 90.5T

degraded zpool Read More »

Scroll to Top