Zagadka's Site

There was a nice suprise waiting at central today on the way into work. 3642 was sitting at platform 1.

I brewed today after a long while. Managed to use up all the old ingredients in two batches so I can start from scratch with fresh yeast and fresh ingredients.

Suprisingly after more than a year of storage, the grain yielded a pretty good amount of sugar and the yeast was actually alive. I was expecting either the yeast be toast or the grain to under-yield.

Cleaned up the fermenter fridge in preparation for the brew day next week!

Came into work this morning to be greeted by OpenNMS sprewing alerts about one of the cPanel machines. Somehow they managed to get over 100,000 emails in the queue. Spam of course.

Today I fitted a new set of bearings to the bench grinder. It runs great now. Very smooth.

Just got home from work. Interesting day - very busy. I changed out one of the core switches today. Very nice piece of kit. The job is a little nerve racking as you must be quite careful of the surrounding connections. Knocking one of those out or kinking a fiber cable and breaking the filaments means very sad times.

I had to push the switch out of the back of the rack, due to patching on the front side. This meant extracting the switch on the front size 80-100mm and then removing the ears on the sides. It was then able to slide to the rear of the rack.

The core switch is a HP Procurve 48 port gig switch. It's got 4 gig fiber uplinks. The other really cool part is the module in the back. The HP J9049. This module allows you two CX4 stacking ports as well as two 10gig GBIC slots.

The lab has expanded somewhat. I now have two layer 3 switches, one with POE for Asterisk work and a new HP 48 port. I also have some loaner firewalls to play with Juniper NS5-GT. I've got VTP going over the cisco switches for vlan management, the HP doesn't talk VTP unfortunately. All of the uplinks, other than to the dumb gig switch are dot1q trunks which carry the vlans between switches.

The Junipers will be serving DHCP and either NATting or Routing for a few vlans which will sit behind the ESXi hosts, more as a learning tool rather than doing anything useful.

I am now hunting down some mini-GBICs for the HP so that I can link the POE switch at gigabit speeds.

Next on the agenda 3 x cisco routers and all the serial crap to go with them.

I managed to track down a newer image with crypto (k) for my Cisco 2950T. So I upgraded it today, rolled the keys and enabled SSH.

I managed to get Asterisk up with FreePBX. Compiled from source - the rpms don't play nice with the current version of FreePBX. It takes a damn long time to build from source on an Atom machine... Next job is to hook up the SPA3102 as an analogue extension and a pstn trunk.

Bought a pair of HP 2650 switches on eBay - very cheap. Unfortunately they weren't packed very well, so the front bezel and GBIC slot on one of the switches are smashed. The seller was very good though. They partially refunded. These are layer 3 capable and should do well sitting behind my ESXi hosts.

I had a bit of a mess around with the new POE switch and the 7940 handsets tonight. I spun up a pre-made Elastix VM which is a CentOS distro with FreePBX and Asterisk integrated - and a bunch of other stuff to make life easier.

I scored a Cisco WS-C3550-24PWR-SMI on eBay.

Once it arrived I gave it a quick test with a Panasonic VOIP handset, the handset was power cycling its self on and off and this was appearing on the console of the switch:

%ILPOWER-3-CONTROLLER_PORT_ERR: Controller port error, Interface Fa0/9 : Power given, but link is not up

I thought to myself, damn, the POE is stuffed on this switch... However after some reading, I discovered that the switch is actually pre-802.3af standard. The IOS image above a certain versino, however, has been fixed so that the switch can provide power to compliant devices. The reason the pana didn't work correctly is because it doesn't boot and bring the link 'up' fast enough for the switch. The switch hits a timeout and turns off the POE on that port. The timeout can be adjusted like this:

power inline delay shutdown 20 initial 300

The switch looks like it's had a hard life though! It came pretty poorly wrapped - but it made it through in-tact. The thing is full of dust.

IMG_20130803_143056.jpg

I scored some second hand Cisco 7940 and 7960 handsets to start experimenting with Asterisk. They look very good for the price.

Had a bit of fun today with my site after discovering nearly 2000 spam comments. Thankfully the comments go through an approval system first before they are visisble.

I also threw django-grappelli on for a pretty admin area. It looks quite neat.

By adding a newer jQuery and jQuery-UI to the admin section I also broke the Zinnia tag autocomplete. I rectified this by replacing the autocomplete_tags.js which utilised Jörn Zaefferer's jQuery plugin with code to call jQuery UI autocomplete.

Some of the brewing gear got justified today in the blackout as we could enjoy hot coffee in the morning with no power.

Oops. This is what the tube looks like if you get the pin mixed up. But be cautious - as this can damage the tube. In this case the decimal point is being used as the anode instead of the actual anode.

Wow, that tube looks awesome running.

I set up the improvised high voltage supply to 250V DC per the data sheet and then dug about for about half an hour to try and find a 22K resistor which was 1W or larger. No joy, but I did manage to find some resistors that match the tube. Some 12K 2% resistors, they look like maybe 2W. Put them in series and we get 24K, or 23.5K in ths case because they aren't in spec (although now that I think of it, they probably are - just, It's about 500ohms out, the 2% gives you 240 ohms leeway each)

With that in place I measured the short circuit current at 8mA. I adjusted it down a bit further to get the current down and ended up with just over 6mA and just over 180V under load.

Next I quadruple checked the datasheet to ensure I had the right pin for the Anode and applied the 180V to the Anode. Then I grounded each Cathode and watched the digit light up. You can see in the photo the number 5 lit.

Wooo! A Nixie Tube. =) Thanks Dave!

This is one of the larger Nixies, it's a GR-211. It has a 170V 5mA strike voltage. We didn't get a chance to try it during the BBQ.

In anticipation of my mate bringing a Nixie tube over to look at I knocked up a quick variable high voltage supply.

It has a massive unloaded voltage of 418V in the photo.

Changed the oil in the car. It's long overdue as you can see by the photo - it doesn't stink yet, but it's pretty dirty.