Archive for April, 2007

This weekend I bought a Linksys SRW2024 switch. It’s a 24 port, 10/100/1000Mbit rack mount switch with a fairly good feature set. The switch it is replacing is a Cisco 2924XL – another feature packed switch – but only 100Mbit. As most things these days come with an onboard Gigabit network adapter, I thought it’d be a nice performance increase to finally get a GigE switch. What I did want however is the switch to support 802.1q trunking. This allows you to use one network card in a machine and have the switch do some magic to make it bridge multiple vlans. This means I only need 1 GigE network card per machine and can have it on 2 logically separate networks. This is a good thing.

Continue reading ‘Review of the Linksys SRW2024 switch.’ »

It’s been a busy month. I’ve finished up with my old employer after my contract ending, started with a new employer and been working flat out. In the past month, we’ve redone a new Xen virtualisation host machine and implemented a new nagios based monitoring system on a DomU guest install of Debian.

On a personal front, I’ve got myself a spiffy new Linksys SRW2024. It’s not a bad 24 port GigE switch. The web interface is a little cludgy, and very hard to use if you don’t use Internet Explorer, and the console interface is pretty much useless. That being said, after it’s been setup, it’s not too bad a switch. For it’s price, it’s pretty good considering it supports 802.1q trunking. I’d love to see Linksys spend some more time on the interface and make the console interface useful (vlan config etc).

Cisco has recently released version 8.2(2)SR1 software for the 79×1 phones (not to be confused with the 79×0 phones!). This seems to fix many issues I was having with the phone randomly crashing, randomly unregistering lines and so on. I’d recommend this software load for EVERYONE running a 7941/61 phone. It is stable, and many issues have been fixed.

The Message Waiting Indicator (MWI) still has issues with a plain asterisk config. I have reported this bug to Cisco and I’m hoping they do actually fix it. In the meanwhile, the smart boys and girls in the Asterisk camp have added a “buggymwi=yes” option that you can add under your phone config in sip.conf to work around buggy Cisco phones. I’m not sure if this affects the 79×0 model phones.

I’m part way through putting Australian tones into the phone – but I’m having issues getting the dialtone to the correct Australian one. The File required is /Australia/g3-tones.xml in Cisco Call Manager – however it isn’t publicly available on the Cisco web site. My current version (with the US dialtone still) is here.

Update 21/09/2007: Thanks to Tim who posted the correct Australian dial tones in the comments, this file is now correct for Australian locale. I have updated the file on the link above. Thanks Tim!

Well, I got suckered in again. I splashed out and purchased one of the Cisco 7961 VoIP phones. As always, it’s very nice gear – that is when it works. Cisco have been very smart with this. They say that their SIP firmware is RFC compliant and should work with any VoIP server – however, sadly, there are that many bugs that stop it from working properly, I don’t even know where to start listing them here.

It’s a really sad day when I can say that my $150 Sipura phone that is 3 years old is more reliable than a $500 Cisco phone. While it does take calls (if you look at it right), the message waiting indicator (MWI) doesn’t work, the phone randomly crashes, the configuration files are not documented at all, the phone randomly drops calls, it randomly unregisters lines, and becomes very unstable after about 12 hours of being powered on – meaning you need to reboot it around twice a day.

To add insult to injury, I have been trying for 3 weeks to purchase a smartnet contract to allow me access to newer software for these phones, however 1) Cisco won’t sell them direct, and 2) none of their resellers are interested in selling me a smartnet contract.

At least the phones look nice – because it still looks like an impressive paperweight.