CBTNuggets

1 03 2012

Looks like I got in there with my CBT purchases at just the right time (I.e the whole of the last 18 months).. Download packages (as opposed to streaming – much less flexible) seem to start from $499 now, rather than the $299 I paid for most of the packages I’ve bought previously ;(

Not saying they’re not worth it, but it’s a big jump.





Instant Messaging

1 03 2012

Have been reminded recently why I shunned IM in the workplace all those years ago; I hate it!

Define IM: a computer application which automatically grants the messenger free reign to break the chain of concentration (however deep the concentration or business critical / time sensitive the nature of the work) of the recipient, to ask trivial questions or attempt to bypass a process, and be entitled to an immediate answer.





Just discovered the throughput limit..

10 02 2012

.. of a Draytek 2910. This was while testing the throughput on a new Virgin Media link. Initially, VM Support told me to set the LAN card on a laptop to 100/full and plug it directly in to the external port of the router. This resulted in a 2Mbps throughput of the link, and triggered VM to come out and adjust light levels.

I thought that the 100/full may be having an impact, so set the card to 1000/full and immediately saw the full 100Mbps obtained from the link. Result.

This was still with the laptop plugged in to the external port of the router, and one of our external IP addresses assigned to the NIC. I added a new “External” VLAN, and plugged both the external port of the router, and the Draytek (which is very much a “standby” bit of kit used for testing – we have a Checkpoint cluster in production), into ports in this VLAN, and I could only achieve 30Mbps.

After verifying that all the port speed and duplex negotiations were correct, I decided to have a dig online – and found several references to the fact that the maximum throughput of the device was 30Mbps (there are a few exceptions – someone has got 40 out of one, but it’s definitely not 100!). The last time I used the Draytek in a testing capacity was when my last fibre connection was installed – but at the time, this was running at 25Mbps, so I wouldn’t have seen the limitation. D’oh!

I’ll know for sure when I carry out the project to make the new connection the primary of two, using our Checkpoint cluster and auto-failover between the two. Gulp.





Busy busy busy

30 12 2011

It’s ironic, and a great shame, that the longest period I’ve gone without posting on this blog has been a period where I’ve had some amazing professional experiences – mostly with stuff breaking that I then had to fix.

The main example of this is the company phone system – we had a complete failure of the OS, caused by underlying hardware issues. The first week of the break I worked 70+ hours to get it sorted, and we still weren’t in a great situation after that! Unfortunately, we went 6 days without the majority of the functionality that the phone system had – and a month later, after much troubleshooting with both our consultants and Cisco TAC, we still didn’t have voicemail.

Now though, I’m confident I could bring the system back to a working state within 4 hours. I have gained lots of experience with the Cisco Unified Communications Manager (or Call Manager, as it was called), both the web interface, and even more significantly, the command line. I have now installed the OS numerous times, so I’m at least familiar with the process. I’d never previously even heard of the Real Time Monitoring Tool, and now I use it every day to keep an eye on the system.

On top of that, I’ve had numerous issues related to my firewalls blocking traffic between my two Exchange DAG members, causing all sorts of fun and games. Incidentally, I’ve just started watching my CBTNuggets 70-662 videos again, and noticed that the guy actually said that DAGs hide the underlying Windows Failover Clustering technology from the administrator.

My troubleshooting during this time has completely disproven this as far as I’m concerned – the Windows Failover Clustering interface has proven invaluable in the course of troubleshooting my DAG, and I always go there first whenever there is an Outlook connectivity issue. It enables you to see at-a-glance which nodes have dropped from the cluster, cluster specific errors, the status of the DAG and the file share witnesses, and to configure new file share witnesses, which doesn’t seem to work properly with Exchange (certainly from the GUI).

Now I’m having intermittent VPN outages that have occurred at almost exactly 12:05 every day for the last five days – this is really heavily intertwined with the firewall and Exchange issues above, so will prove an interesting one to troubleshoot.

I’m aiming to write these experiences up in more detail soon – I documented my troubleshooting quite extensively – back soon!

PS – hope you all had a good Chrimbo!