Telecoms overload

I feel like I’ve got “telecoms overload” today.  I’ve just created a new virtual machine to hold the latest Asterisk (*) 1.8rc2 release.  That will bring the total number of PBXes running in this flat to three.  I’m also threatening to, if I ever find the time, to see if I can figure out one of the SER family (e.g. Kamailio, OpenSER, OpenSIPS etc.) just for good measure.  All I can say is I’m glad I don’t have to finance a real physical machine for each server – virtual machines were a great invention!

I’m still no further forward in deciding what I’m going to about my “telecoms infrastructure” (if you can call it that, since it sounds so pretentious).  There’s now so many products to choose from, and each has their advantages and disadvantages.  Asterisk, the open-source PBX that started it all, has nearly reached version 1.8 but I’m fast falling out with it on the basis that its SIP module is still far from RFC compliant.  FreeSWITCH, on the other hand, is a much more modular, clean design, and a standards-compliant SIP stack, but the XML configuration files drive me up the wall.  The OpenSER/SIPS/Kamailio family are also very capable and powerful softswitches, but you need a degree in RFC 3261 just to configure them!

My head hurts…

Birthday curry

Had a very pleasant evening last night with 12 other people celebrating my forthcoming birthday tomorrow in a nice little restaurant about five miles away from here.  The chicken madras I had was pleasantly on the hot side, and there was plenty to go round (in fact I think I was one of the few people to actually finish my curry!).  Highlight of the evening was the person who shall remain nameless – but you know who you are! – who phoned me up to ask where we all were.  I told him he was in the wrong restaurant, five miles the other way, to the amusement of everyone else on the table :)

Tired of tyres

A foot pump is a foot pump is a foot pump, yes?  Well, no.  It would appear that foot pumps (and especially their pressure gauges) are not necessarily as accurate as you might think.  About a month and a bit ago, my “trusty Halfords foot pump” decided to split in half.  Fair enough – time for a new one.  Nearest place to get one was Tesco (a good few miles away from where I was at the time), so off I go and buy a new foot pump.  “That’s odd”, I thought, when attaching the new pump on the car tyre to check its pressure.  The gauge, compared to the Halfords one, was overreading by some 0.4 bar (5.8 psi).  So which one was right?  The answer, of course, was that I didn’t know.  Given that I was going to be driving back some 400 miles a few days later, I was a bit concerned, but there was nothing I could do about it, so when I got back home I decided to do a bit of research.

The first thing to do was to get the trusty old Woolworths (remember them?) digital tyre pressure gauge out, just as a quick check.  And sure enough, that was reading quite a bit lower than I thought.  This got me wondering if there were actually some standards for tyre pressure gauges.  A quick search on the Internet revealed that there was – and it’s called BS EN 12645:1999.  EU Directive 86/217/EEC states that for gauges that measure  tyre pressures up to 4.0 bar (58 psi), the maximum error must be no more than 0.08 bar (1.16 psi).

So it appeared my not-very-trusty Tesco foot pump was reading outside that range, so the hunt was on to find a pump, foot or electric, that was at least that accurate.  I tried looking through some reviews on the Internet.  It appears that some foot pumps were very very off in their readings; one review of a Draper footpump claimed it was 5 psi out compared to their reference gauge.  So I quickly came to the conclusion that all footpumps were rubbish, except maybe the Michelin which according to the reviews had a very accurate gauge – the only trouble was that many of the reviews said that despite the 30 quid price tag the pump fell to bits very quickly.

So who makes accurate gauges?  Well, my first stop was to the company that produces many of the ‘pro’ tyre pressure gauges you find in garages, tyre fitters, etc.  That company is called PCL, based in Sheffield.  The one you’ve probably heard of and may well have used at a petrol station is their Mk3 pressure gauge.  This is not only BS EN 12645:1999 compliant but comes with a calibration certificate to prove it too!  OK, so that’s got to be accurate, but no doubt dead expensive.  Although not as expensive as I thought… ££ rather than £££.  And anyway, that’d be no good since I would actually need to connect it to an airline.

A friend of mine suggested the Ring Automotive pumps, which are available in most of the usual suspects (Halfords et al), but I wasn’t sure how accurate it was – the web site certainly wasn’t going to mention it.  So my friend ended up emailing the company, and they replied it was accurate to +/- 5%.  This equates to a max error of 0.11 bar (or 1.58 psi) – which is still worse than the EU directive.

Could I find anything better?  Well, it appears it’s PCL to the rescue.  As well as making tyre pressure gauges that cost £££s, they also do a range of cheaper ones, and also an electric pump model.  Their manual gauges are actually calibrated to BS EN 12645:1999 and cost about a fiver, but of course that’s only a gauge.

But…. hidden deep in their web site is an apparently little-known product that they do called the ATT741.  It’s a “consumer” tyre inflator in that it comes in a box, with a cigarette lighter plug on one and and the thing to put on your tyres (no idea what that’s called in reality) on the other, and a box in the middle with a digital display.  The good news is that the web site claims it has a reading accuracy of 1 psi.  That’s 0.069 bar, which is actually better than the EU directive.  Bingo! I might have found one at last!  So I found a supplier on the Internet and ordered it over the weekend.

Today, it arrived.  (For some reason a couple of my colleagues at work wanted a go on it as well when they discovered what it was… which was just as well because they both had quite underinflated tyres as it turned out…)  So, it’s been given a good test.  Reading the instruction book, the manual claims it’s the reading accuracy is 5 kPa, or 0.05 bar, which in English money is 0.72 psi, actually better than the 1 psi the web site claims.  No way to prove this of course, since there is no BS EN 12645:1999 calibration, but I’m sure it can’t be that bad, after all PCL are used to making accurate gauges.

So, I’ve used it on my tyres now, which were quite a bit underinflated as well it turns out, and the car certainly handles a lot better now.  And at least I can be reasonably confident that the real tyre pressure is something close to what the gauge says it is.  Although, if I’m feeling particularly paranoid, I could always check it against a BS EN 12645:1999 calibrated gauge (!)

Details of the PCL ATT741 can be found on the PCL web site.  And yes, the Tesco foot pump is going back to Tesco.

Kernel exploits

Big panic in the Linux world after an article on Slashdot here alerted me to the fact that a 2-year old exploit had just been found in the Linux kernel. This only affects 64-bit (amd64) kernels, but the trouble is that all of my machines and virtual machines run a 64-bit kernel. Panic!! So, I spent a little while last night upgrading about 8 kernels, then rebooting them all, then finding out they wouldn’t start up again properly because of an error in another script, fix that…. all a bit painful. But it’s done now, and hopefully that’ll be the last emergency kernel patch I’ll need to for a while.