[ Home ]      [ Blog ]      [ Tutorials ]      [ Aviation ]      [ Manuals ]      [ Grad School ]
 

Sunday, August 19, 2007

A Good Time Was Had By All

This weekend we saw Becky's family for a big family dinner on Sunday. Lots of folks were there and there was LOTS of good food. Becky's brother, Greg, and I spent a little while digging through the depths of the barn. We found lots of cool old stuff: wooden crates for tank ammunition, lots of Lionel train bits and some 15 year-old hard cider. The bottles of cider were still sealed, so we brought a few bottles home with us to try later. I wonder if it'll be any good...well aged cider can't be too bad. :)


Becky's cousin Cori was running around and pretending to be a cat for most of the day, which made for some fun pictures. Here she is being silly:

We were back home by 9 o'clock. Which is good, because I'm on-call for rescue squad tonight. Tomorrow will be a busy day at work, so hopefully it is a quiet night.

Tuesday, July 03, 2007

An Early Morning

Time: 0530
Today was an early morning. I had an outage scheduled to begin replacing a core switch in one of our academic buildings. Unfortunately, this promises to be a labor intensive outage, since I am replacing a non-trunking switch with a newer model (that does trunking). The ability to trunk is a requirement for the ongoing VoIP telephone project (and this should have been done a long time ago). Because none of the edge switches in this building are currently configured for trunking, I have to patch a closet into the new switch and then go to that closet to configure trunking and VTP. Then I return to the core switch closet and patch another closet, then go to it and configure trunking and VTP. Repeat ad nauseum (or 11 times, which ever comes first). I finished my fourth closet just before 0800 and decided that was a good stopping point (since users begin arriving around that time). I'll need to come in early on Thursday (since I hope to be off on July 4) to finish up the task.

Time: 0745
Shortly after stopping work on the core switch replacement I got a phone call from a vendor who was on campus to install a pair of T1s for the VoIP project. They needed to be let into a switch closet in an academic building to patch their fiber through to our Network Operations Center (NOC). Then we spent some time bringing up the electronics in the NOC to terminate the T1s. After some brief excitement with a broken fiber patch, both T1s came up and tested good from the central office (CO). This is a very good thing, because I don't feel like tracing broken fiber. The vendor left to go and repeat the process for the downtown campus.

Time: 0930
I finally get a chance to sit down and check my email. I have 60 new messages since yesterday evening. About 20 of these are SPAM and get deleted right away. Another 20 or 30 are automated messages alerting me to some impending danger or horribly inappropriate behavior on a system. These emails are skimmed and ones that tell me about something particularly nasty lead to some brief diversions (either to fix the problem or to send a heads-up email to someone who can). The remaining messages, usually about ten in number, are emails from users regarding existing issues. This means that I need to open up our ticketing software to record my responses and close tickets. I always hope to closet alot of tickets (or to have no tickets to close)...rarely does this happen.

Time: 1100
I find a pile of forms on my desk from yesterday's work in the closet. I pull together these forms and put the information that they represent into our telephone tracking database. This is a tedious process, but requires a knowledge of the network and the (new) VoIP telephone system...so I get to do it.

Time: 1215
Ron pokes his head in my office and asks me if I want to do something to lunch. We end up getting take-out food and I come back to my office to eat it. I need to get the rest of these forms processed and into the database.

Time: 1330
I notice some students sitting around the HelpDesk, so I put one of them to work finishing up the cabling the new server rack. I started this cabling last week, but never was able to get back to it to finish. At the time, it was the perfect Friday task: very little thought required, work in the NOC on a hot day. You get the picture.

Time: 1415
Its time to get out in the field with the crews of guys that are inventorying and cabling for the VoIP project. This is a thankless task (have you seen a pattern yet?) that usually needs to happen in a cramped, hot and humid closet. Did I mention that it doesn't go quickly.

Time: 1545
My crew finished one of the larger closets in the same amount of time that it took the other guys to do two slightly smaller closets. I've decided that I'm slow at this process. Others might suggest that I just need more practice. I gather the data sheets from the other crew and start inputting them into the database.

Time: 1630
It's time to go home. Tomorrow is a holiday and I've reached a decent stopping point. With any luck, things won't go up in flames tomorrow. Just enough time to solve the mystery of the "strange disk quota" on my way out.

A Year in the Life of a Sysadmin

So I have decided to keep a record of my life as a systems administrator. I'll post a log entry here for each day that I do work, including weekends and holidays. Some will be short and some will be longer. If it is a busy day in the office, this may turn into a twitter-ish blog. It is my hope that this will provide insight into the life of a system administrator (and it will let my wife know what I do at work too).

Thursday, March 15, 2007

Burning DVDs without cdrecord

I learned today that the DVD burner on my IBM T60p is not recognized by cdrecord, which makes it difficult to burn DVD ISOs to DVD. Or so I thought...

I was able to grab the /dev designator for my DVD drive by doing:
dmesg | grep DVD

This told me that my DVD drive was /dev/hda. Then I was able to use growisofs to write the ISO to DVD:
growisofs -Z /dev/hda=/path/to/dvd.iso


It was just that simple.

Burning DVDs without cdrecord

I learned today that the DVD burner on my IBM T60p is not recognized by cdrecord, which makes it difficult to burn DVD ISOs to DVD. Or so I thought...

I was able to grab the /dev designator for my DVD drive by doing:
dmesg | grep DVD

This told me that my DVD drive was /dev/hda. Then I was able to use growisofs to write the ISO to DVD:
growisofs -Z /dev/hda=/path/to/dvd.iso


It was just that simple.

Tuesday, September 12, 2006

A little bit of sad

Today marks the end of an era. We took down a system for trade-in today. One of the last commands that we issued on the system was a "sh ver". It reported an uptime of 913 days, 21 hours, 11 minutes.

Tuesday, August 01, 2006

Security Methods to Consider

The following security considerations should be employed/considered:

  • Ingress/Egress filtering at VLAN edge - This will prevent a malicious user from spoofing their source address to gain access to restricted services.

  • Separate departments using VLANs - We should consider dividing the adm network into departmental groupings

Sunday, May 14, 2006

Linux search and replace in multiple files

So, I needed a little tool to perform search and replace on multiple files in a Linux environment. Here's how I did it:


#!/bin/sh
for file in *; do
mv $file $file.old
sed 's/FINDSTRING/REPLACESTRING/g' $file.old > $file
rm -f $file.old
done

Tuesday, April 18, 2006

Fedora Core 5 on an IBM (lenovo) T60p

So, I just got my new IBM (lenovo) T60p laptop. It came preloaded with a bloated, vendor-hacked version of Windows, which I couldn't bear to leave on there. So, I busted out my FC5 CDs (all five of them) and got my install on. There were a number of devices which were must-haves from my perspective:

  • Video: I had to have a good display revolution, but running ATI's drivers were not high on my requirement list

  • Network: Both the Intel Pro 1000 (eth0) and the Intel Pro Wireless 9345 A/B/G wireless (eth1) had to be fully functional

  • CPU: This beauty has the new Intel Duo processor, which means we get to run an smp kernel! WOOT!



So, here's how I got those guys working:

Video - This worked "out of the box." It doesn't use ATI's driver yet, but I get good video quality using the vesa driver, so I probably won't try to change unless I have a compelling reason.


Network - The wired interface worked right out of the box. The wireless was a pain in the /dev/null. The FC5 kernel comes with ieee80211 compiled in. We shouldn't mess with this, we will only mess it up. Trust me. After I installed, I 'yum update'ed which brought my kernel up to 2.6.16-1.2080_FC5smp. Luckily, I found an atrpms mirror that had RPMs for the ieee80211-kmdl and the ipw3945* items. So I grabbed and installed these rpms:


    ipw3945d-1.7.18-1.at
    ipw3945-ucode-1.13-1.at
    ipw3945-kmdl-2.6.16-1.2080_FC5smp-0.0.74-4.rhfc5.at
    ipw3945-0.0.74-4.rhfc5.at

Reboot and then run /sbin/ipw3945d. Now you should see a new interface (eth1 on my box). Viola! You can use your iwconfig tools to manage it just like it was a regular network interface.


On Screen Display -
http://www.nongnu.org/tpb/

   ©Copyright 2004. All Rights Reserved