Sles10

Disclaimer: I don’t take any responsibility for faults within the software, I just provide the RPM’s! Feel free to ask me about stuff concerning these RPM’s, but I ain’t accountable if your stuff goes kaboom … Oh, and those RPM’s aren’t recommended or supported by Novell or IBM! After working with the novell-kmp solution, I think it’s actually rather easy to create a " Kernel Module Package". In the end, I created two additional KMP’s, one for the tools component of the VMware-Tools shipped with VMware ESX, and another for the lin_tape SCSI driver, used by our IBM TS3400 as well as the IBM TS7530. ...

Well, I just stumbled upon something .. My Nagios at work wasn’t working anymore, and I went looking. 1 2 3 4 5 6 7 8 9 10 11 nagios3 ~ [0] > tail -f /var/log/nagios/nagios.log [1238658394] Error: Unable to save status file: No space left on device [1238658403] Error: Unable to save status file: No space left on device [1238658413] Error: Unable to save status file: No space left on device [1238658423] SERVICE ALERT: tsm1;POWER WARN;OK;SOFT;4;-u OK - 0 [1238658423] Error: Unable to save status file: No space left on device [1238658433] SERVICE ALERT: tsm2;LOAD;WARNING;SOFT;1;WARNING - load average: 6.25, 5.72, 5.36 [1238658433] Error: Unable to save status file: No space left on device [1238658443] Error: Unable to save status file: No space left on device [1238658453] Error: Unable to save status file: No space left on device [1238658463] Error: Unable After that, zip - nada. Next thing, check whether or not the device is really full … Okay, df .. ...

In the past, I always had problems with SLES and our Tivoli Storage Manager client’s when backing up files with german umlauts. Well, today I looked a bit harder, and quite quickly found a solution. 1 2 sles9 root [0] > env | grep ^LC LC_CTYPE=de_DE.UTF-8 As you can see from the above, SLES9/10 ain’t setting LANG or LC_ALL (which I searched for first), but is setting LC_CTYPE. So, simply changing the LC_CTYPE in the init-script and/or prepending the dsmc command line with a new LC_CTYPE fixes my umlauts problems! ...

In our current fight against the BladeCenter switches, we’re currently facing the problem that the blades ain’t able to send/receive DHCP-traffic. So in order to move forward, we had to use static IP addresses. And since SLES10 ain’t straight forward on that, I had to look it up. Now, here’s for me (and everyone else tired of searching) how to do it: 1 2 3 4 5 6 7 8 9 install=http://ftp.barfoo.org/install/SLES10-x64/CD1 netdevice=eth0 hostip=10.0.1.240 netmask=255.255.255.0 gateway=10.0.1.1 nameserver=10.0.1.2 domain=home.barfoo.org insmod=bnx2 splash=verbose

Well, I recently (well, yesterday) built the opsview RPM’s for SLES10, and started fiddeling about with it today. Alex " recommended" I should rather look at Opsview instead of Centreon, but boy was there a surprise waiting for me … Opsview has the advantage that it at least lets you use the package manager. But, it also needs a lot of handy work (just like Centreon, which I really dislike since it’s real error prone). ...

I had the problem, that the automatic update function of YaST doesn’t work like I want it to. I just wanted it to install only those updates, that ain’t interactive, don’t need a service restart and don’t need a reboot. YaST does only feature an online update that skips “interactive” updates (I’ve never even encountered an interactive update up till now). So I went ahead and wrote a (hackish) script, that achieves what I need. ...

The other day I had a closer look at the zypper logs (well, I was digging for a time-history of installed packages). First … damn does zypper produce a lot of logs on a " productive" (or rather on a maintained - as in up-to-date) system. But glazing over the logs, I found out something new about zypper. It actually has an internal list, which only purpose is to identify a trusted vendor … ...

Well, I just looked into using ` zypper up ` to update some of our boxen (I do have a script, which holds the boxen it needs to process in a variable and simply goes through them one by one) – yes, I could activate auto-update, I just don’t want that at this point 😉 So at first I tried just using zypper to automatically update that given list, but even if you pass –no-confirm, zypper would still ask for your confirmation (which seems kinda stupid). After a short while thinking about it, a lesson from solar@gentoo.org came to mind. When working in a chroot, he simply used this: ...

Well, I was just a bit curious earlier what distribution might be running on our IBM TS7530 Virtualization engines .. well, I just had a look-see .. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 vetapeservice@VTL-B:~> cat /etc/SuSE-release SUSE Linux Enterprise Server 10 (x86_64) VERSION = 10 PATCHLEVEL=1 vetapeservice@VTL-B:~> uname -a Linux VTL 2.6.16.46-229-smp #1 Sun Apr 13 05:21:49 UTC 2008 x86_64 GNU/Linux vetapeservice@VTL-B:~> free -m total used free shared buffers cached Mem: 4022 1378 2643 0 342 552 -/+ buffers/cache: 484 3537 Swap: 3815 0 3815 vetapeservice@VTL-B:~> cat /proc/cpuinfo processor : 1 vendor_id : AuthenticAMD cpu family : 15 model : 65 model name : Dual-Core AMD Opteron(tm) Processor 8218 stepping : 3 cpu MHz : 2600.186 cache size : 1024 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 Main difference to a " normal" SUSE Linux Enterprise Server 10 installation (there’s about zip normal with that kind of installation, thus the quotes) thus far are: the build for the VE uses busybox as init IBM stripped man/info they are running Xorg/Fluxbox on it Just don’t ask me why there’s a DE ( d esktop e nvironment) running, it ain’t even hooked up to a monitor. Only reason would be for the RSA’ remote monitor stuff … lala

Well, today I had a rather weird error. I was testing the adapter bonding on one of the boxen designated as Tivoli Storage Manager Server, when I noticed that the bonding wasn’t working as expected when simulating an error (that is unplugging one of the TP cables for the bond). Now, the bond had “mode=6 miimon=100” as options. After running “linux bond debug” through Google (which turned up nothing useful, besides one document on the Oracle Wiki about IOS/Linux adapter teaming), I figured “Hey, just lets test switching the arguments.” And guess what ? ...