Jonathan Oxer
[Blog]
>> OpenQRM
Tue, Aug 29th 4:03pm 2006 >> Linux
Chris, my friend, you are a champion.
I'd never heard of the OpenQRM project but it looks very cool. I'll have a play with it when I get over the stress of trying to get this Xen cluster running properly! I'm still having big issues with the storage system because I want redundant storage, and all the howtos I've managed to find talk about just having a single storage server. Lots of scribbled diagrams, lots of head scratching, but no neat solution yet.
Chris, my friend, you are a champion.
I'd never heard of the OpenQRM project but it looks very cool. I'll have a play with it when I get over the stress of trying to get this Xen cluster running properly! I'm still having big issues with the storage system because I want redundant storage, and all the howtos I've managed to find talk about just having a single storage server. Lots of scribbled diagrams, lots of head scratching, but no neat solution yet.
>> SLED/SLES 10 launch
Wed, Aug 23rd 5:47pm 2006 >> Linux
Went to the Suse Linux Enterprise [Desktop | Server] 10 launch bash put on in Melbourne this morning by Novell, and I've got to admit that it's damn impressive. Most of my time over the last month has been setting up Xen-based virtualization for IVT's new hosting infrastructure at Verizon, and I've done it all with Ubuntu which has lead to some pain when it comes to the storage side of things. I'm running VM images stored on an NFS server, with the Xen hosts mounting an export with all the VM roots in it. Works fine but took a bit of work to get going, and it makes things like storage failover quite hard to manage.
The work Novell have put into SLES makes all that stuff a piece of cake. Not only do they ship Xen kernels (which a few people are doing now) but they've gone the extra mile and included really polished support for iSCSI, heartbeat, DRDB, and all the other little bits you need to build a proper HA system. Xen on its own is all fine and well when you're running a couple of VMs on one box, but when you're building a server farm with a bunch of front-end machines, multiple storage machines, multipath routing, and automatic failover it's a whole different story. Seeing how easy SLES 10 makes it was a real eye opener.
Went to the Suse Linux Enterprise [Desktop | Server] 10 launch bash put on in Melbourne this morning by Novell, and I've got to admit that it's damn impressive. Most of my time over the last month has been setting up Xen-based virtualization for IVT's new hosting infrastructure at Verizon, and I've done it all with Ubuntu which has lead to some pain when it comes to the storage side of things. I'm running VM images stored on an NFS server, with the Xen hosts mounting an export with all the VM roots in it. Works fine but took a bit of work to get going, and it makes things like storage failover quite hard to manage.
The work Novell have put into SLES makes all that stuff a piece of cake. Not only do they ship Xen kernels (which a few people are doing now) but they've gone the extra mile and included really polished support for iSCSI, heartbeat, DRDB, and all the other little bits you need to build a proper HA system. Xen on its own is all fine and well when you're running a couple of VMs on one box, but when you're building a server farm with a bunch of front-end machines, multiple storage machines, multipath routing, and automatic failover it's a whole different story. Seeing how easy SLES 10 makes it was a real eye opener.
>> Amex sucks even harder
Wed, Aug 23rd 5:27pm 2006 >> Bad People
After my previous bad experience of Amex suspending my card while I was trying to pay for my travel to OSCON I got a bit paranoid about it, so after they reactivated the card I paid an extra $1000 onto it even though the next bill hadn't even been sent to me yet. AFAIK it's not possible to have a positive account balance on Amex: you spend the money, they send you a statement, and you have a month to pay it. But I didn't want to end up in the US desperately needing the card only to find out they'd sent out a statement while I was gone and suspended the card again, so the extra grand was a show of good faith that the next bill would be paid.
So what happens? In the US I arrive at the hotel in Portland, try to pay on Amex, and it's declined. Again. *After* I've sent them an extra $1k. *Before* the next statement is due.
I think I've set a new record. Not only have I had an Amex card suspended when my account was fully paid up, but I've now had one suspended after I had paid them more than they'd asked me for.
My reaction last time was "screw you, Amex".
That doesn't even start to cover it this time.
After my previous bad experience of Amex suspending my card while I was trying to pay for my travel to OSCON I got a bit paranoid about it, so after they reactivated the card I paid an extra $1000 onto it even though the next bill hadn't even been sent to me yet. AFAIK it's not possible to have a positive account balance on Amex: you spend the money, they send you a statement, and you have a month to pay it. But I didn't want to end up in the US desperately needing the card only to find out they'd sent out a statement while I was gone and suspended the card again, so the extra grand was a show of good faith that the next bill would be paid.
So what happens? In the US I arrive at the hotel in Portland, try to pay on Amex, and it's declined. Again. *After* I've sent them an extra $1k. *Before* the next statement is due.
I think I've set a new record. Not only have I had an Amex card suspended when my account was fully paid up, but I've now had one suspended after I had paid them more than they'd asked me for.
My reaction last time was "screw you, Amex".
That doesn't even start to cover it this time.
>> Stop RFID
Mon, Aug 14th 11:15pm 2006 >> Tech Toys
One of the people I was lucky enough to meet at OSCON was legendary PHP hacker Zak Greant, who had lunch with Arjen and myself after the conference close on Friday. I'd just done my "Making Things Move" talk which included a demo of reading the RFID tag in my arm using common scripting languages, and Zak had a prototype of a groovy little RFID detector badge produced by German group FoeBuD so we quickly got to talking about RFID and privacy.
FoeBuD are strongly opposed to the proliferation of RFID due to the privacy issues, so they've created an impressively simple RFID detector that consists of a sum total of 3 surface-mount parts (plus the coil itself on the PCB makes 4) and which lights up to warn you when you're being scanned by an RFID reader. The idea is to wear it around and see how often you're being subjected to scans in everyday situations such as shopping centers, petrol stations, etc, and since the badge happens to be bright red and say "STOP RFID" in huge letters it's a bit of an attention-grabber too.
So while we were sitting at lunch Zak and I tried to use one of my home-made RFID readers to trigger the detector, but without success. I assume the detector is designed to operate with the high-frequency (HF) tags typically used in supply-chain systems, while my reader is a 134.2KHz low-frequency device for reading implantable tags. Damn. Zak was kind enough to give me the prototype though, so I'll have to take it shopping and see if I can find a store that uses readers that trigger it.
One of the people I was lucky enough to meet at OSCON was legendary PHP hacker Zak Greant, who had lunch with Arjen and myself after the conference close on Friday. I'd just done my "Making Things Move" talk which included a demo of reading the RFID tag in my arm using common scripting languages, and Zak had a prototype of a groovy little RFID detector badge produced by German group FoeBuD so we quickly got to talking about RFID and privacy.
FoeBuD are strongly opposed to the proliferation of RFID due to the privacy issues, so they've created an impressively simple RFID detector that consists of a sum total of 3 surface-mount parts (plus the coil itself on the PCB makes 4) and which lights up to warn you when you're being scanned by an RFID reader. The idea is to wear it around and see how often you're being subjected to scans in everyday situations such as shopping centers, petrol stations, etc, and since the badge happens to be bright red and say "STOP RFID" in huge letters it's a bit of an attention-grabber too.
So while we were sitting at lunch Zak and I tried to use one of my home-made RFID readers to trigger the detector, but without success. I assume the detector is designed to operate with the high-frequency (HF) tags typically used in supply-chain systems, while my reader is a 134.2KHz low-frequency device for reading implantable tags. Damn. Zak was kind enough to give me the prototype though, so I'll have to take it shopping and see if I can find a store that uses readers that trigger it.
>> OSCON: the wrapup
Mon, Aug 14th 11:02pm 2006 >> Conferences
OSCON was fantastic, pity I didn't get to see it! Or not much, anyway. Damian Conway's speaker training class on the Sunday night before the first tutorial day was great, but it meant that I frantically spent the rest of the week restructuring my talks based on his advice so I only ended up going to about 4 talks other than my own :-(
The extra preparation was worth it in the end though. My last talk (the hardware hacking one) got a fairly serious makeover in terms of structure and flow compared to how I presented it at OSDC last year, and it ended up flowing much better. I had quite a few people tell me afterwards it was the best session of the conf, which was pretty amazing and a huge compliment given the company I was in! I know I'll never be even in the same ballgame as Damian but at least I can have fun trying.
After flying back in and arriving late morning on the Monday I was back working within a few hours and haven't had time to stop since.
OSCON was fantastic, pity I didn't get to see it! Or not much, anyway. Damian Conway's speaker training class on the Sunday night before the first tutorial day was great, but it meant that I frantically spent the rest of the week restructuring my talks based on his advice so I only ended up going to about 4 talks other than my own :-(
The extra preparation was worth it in the end though. My last talk (the hardware hacking one) got a fairly serious makeover in terms of structure and flow compared to how I presented it at OSDC last year, and it ended up flowing much better. I had quite a few people tell me afterwards it was the best session of the conf, which was pretty amazing and a huge compliment given the company I was in! I know I'll never be even in the same ballgame as Damian but at least I can have fun trying.
After flying back in and arriving late morning on the Monday I was back working within a few hours and haven't had time to stop since.
1 of 1
[ Back to top ]