Older blog entries for rince (starting at number 11)

I never would have dreamed that my weekend can be such a nightmare. The official plan was to have a 3 hour downtime to upgrade the firmware on the switches. I said that's fine by me. I will have a visitor but for three hours that's okay.

After the upgrade the nightmare began: the cluster wasn't able to start again. Unfortunately, the distributor wasn't able to find the problem... at least until Sunday evening. So we had around 25h downtime of a productive cluster, my visitor had to look out in Stuttgart for himself and I had two days in the bureau where we just had to wait...

at least everything works fine again.

What a start in the day... I leave my appartement and my mobile rings - a collegue calls, one node of the cluster is down. So much for walking into the bureau; now I have to take the tram to help him. Oh, well...

Do you know that Solaris-Updates can be quite difficult? I do not really like it, but I have to make them...

Well - it has been now 10 months without any diary entry from me. A lot of things have changed since then. First, my former company got bancrupted. KPNQwest Germany broke down, my project has been taken over by the company which did the marketing for the project before (and since a few days _all_ former technicians are gone from that project, we'll see how long it will survive :-) I myself found a new job - not in Karlsruhe anymore, but in Stuttgart, a town nearly 80 Kilometers away. Everyday I fetch the train now to get in there. What I do there? Mostly operations - running a realtime-clustersystem. Quite some open source is used, but not enough - if I am able, I'll try to bring more in, like netsaint for monitoring. Let's see what I can do here.

Hi guys, welcome to my last diary-entry before Christmas.

Last night I was in Lord of the Rings. To be honest: it was not as good as I hoped. Too much playing around with music or Sound Effects, the fights were quite chaotic to follow and part of the books are rewritten. For the rest: it's fun to see, but don't try to be a fan or a guy who looks for every word when you watch it.

Strange things happen again...

I just try the new Opera (5.05 under Linux, after 6.0 crashed immediately) and have to find out that I am on Master level. Funny thing, ey?

Remember my last entry? About the customer project? Well - the project went more or less fine, now we have fun with the continuity of that project; if it goes the way I want to have it I need more people working here (customer wanst 24/7 and I refuse to make this with three people everytime). So I will see what will come over the jump to the next year. Btw, we replaced our NFS System with a new one from NetApp. That happened September. Since then we never had a problem anymore with the system - that's what I wanted to achieve a long time before but wasn't allowed to :-)

Listening to PDQ Bach is quite fun, especially at work. If you have people working around you and you listens to Classical Rap, they get nearly furious *grin*

Strange things happen...

I tried to leave our most public projects because I disliked the way of disinformation within the company. Now I became teamleader. Especially nice because I now see that communication has been improved.

Anyway - it was a good start in the new year. Our NFS- System is fine now as long as no one (also not the vendor) touches it.

Today I discovered that IceCast is a nice way to hear the mp3s from home in my office - it works quite fine if you have a good downlink :-) Thanks to debian/NetBSD, it worked quite fast so I was easy starting all the stuff.

This morning, we want to make the final clustertests for a new projekt. First we had to find out why Solaris doesn't want to work with Cisco Switches - since Solaris 8 they have changed the behaviour which is quite surprisingly. Well - until midday we finished reconfiguration of everything....

Now we try to test the cluster. I think it is a lot of fun to crash a machine by purpose - just to do it makes enough fun because you expect it to come back. If it doesn't, blame your consultant who installed the whole stuff :-)

I really like our black-box-NFS System. We told the vendor that they have configured wrong the network-components within their system, but they wanted to believe until we proved it to them. Then they had to reconfigure every one of their datamover. Guess what they forgot: The Hotspare-Datamover. It was not really nice to see that there was a failover (which is okay if it works) - and that this failover has not resolved the problem. Oh, btw: the second path of alerting people will soon be email after the modem was not able multiple times to call out... anyway, who cares? At least I know now what I have to know about big storages :-) And I love to have ConCalls just because of these accidents.. What else? Has anyone tried to install Solaris 8 via CD? Whoever had this idea with using 5 CDs or more just to install a plain Solaris must be mad - it takes sooo long time... Where is my jumpstart-server-on-the-laptop...

This is a good day. This night we had a routing problem in the backbone - which meant that the servers I am working with were not reachable. Which meant that I got alerted.... Well - because it was a routing problem, I haven't had much to do except of traceroutes. After one hour that problem was fixed.

Then I had to make measurements how fast our black-box-NFS-Storage can be, mounted via NFS to a Solaris box. Its interesting; if the datamover is really idle you can find >8000 directories with "find /mount/point" within 10 seconds. But you have a datamover who is working this drops down very far....

And I just got the recent news: our company changed the mobile provider. Three days ago. Now everyone who has a company mobile has to switch to the new provider (now!), use the new number, copy all his numbers in the memory to the new card, publish the new card......

I like chaos, especially when I am not involved (my mobile is a private one).

And btw: the Apocalyptica Concert in Karlsruhe next week is sold out. Too bad.

Have I told you that I like environmental monitoring made by other departments? I like it to hear in the morning that there was a power peak (I knew this one before because my machines at home booted suddenly in the night, so much for a nice uptime) and now one really cared about the powerpeak in the computer rooms - well, someone discovered that there was a problem somewhere but who cares...

Anyway, I am surprised to find out that nearly everything is running fine, so I can concentrate on documentation and development - finding out how cfengine really works and how to get it used with checksums for some files :-)

It is a new week. It started with reading the mails of the last three days for work (including power loss in another city, routing problems and so on). Last week was surprisingly quiet - no real panics, no deadlines (yet). But that was only three real days.

Last weekend I examinde the "Citynightline" in Germany - enter the train at 23:20 in one City and arrived at 07:30 in another one. Nice trains, nice offer for a good trip. It was fun meeting some guys out there without having to wait to arrive the town.