Saturday, June 21, 2008
Failing and webcasting |2:55 AM| Suzy's dedication and commitment of force is admirable.
I help do the webcast for a local sporting event. Problem is, my tech level is a bit limited on the server side. I rely on a different company for my bandwidth, and I only have one encoding box.
This means, basically, zero redundancy. If something goes wrong, the whole audience sees and I look like an asshole.
Later today I'm going to try to simulate a total failure of the transcoding box, (like some passerby tripping over a cable or the firewire card buying the farm, or spontaneous combustion) while I switch over to the laptop as a temporary server. This is a less than ideal solution as it complicates matters while not addressing the vulnerability of my setup to such failures. Also, few advanced technologies have backup systems that can be described as "Switching cables really fast and hoping you can click the mouse at the right moment". I've set the system up so I have 29 seconds to perform this trick before anyone notices.
I hate not having a 99% success rate on this, this is known technology. Instead, I average about one reboot per match. If this was a Blackberry Enterprise server I was running as my job, I'd likely have already been fired.
Our setup, our system, is all distinctly home brew. Hell, previously, I was running this on a Ubuntu box I learned to use in less than a week. The library lacks books on this specific subject, so I'm trying to learn from online sources. Apparently the big boys (like say, the local massive college sports department) have magic, $8000+ boxes that do their webcasting* by just being plugged in. That's called a "turn key" solution, is outside my price range, and anathema to my need to understand what's going on. If my fellow crew members want to take this show on the road, I've got to put this all together myself, and properly.
*(Or they're supposed to, I'm told by an informed source that sometimes they don't work and no one knows how/why)
Failing and webcasting |2:55 AM| Suzy's dedication and commitment of force is admirable.
I help do the webcast for a local sporting event. Problem is, my tech level is a bit limited on the server side. I rely on a different company for my bandwidth, and I only have one encoding box.
This means, basically, zero redundancy. If something goes wrong, the whole audience sees and I look like an asshole.
Later today I'm going to try to simulate a total failure of the transcoding box, (like some passerby tripping over a cable or the firewire card buying the farm, or spontaneous combustion) while I switch over to the laptop as a temporary server. This is a less than ideal solution as it complicates matters while not addressing the vulnerability of my setup to such failures. Also, few advanced technologies have backup systems that can be described as "Switching cables really fast and hoping you can click the mouse at the right moment". I've set the system up so I have 29 seconds to perform this trick before anyone notices.
I hate not having a 99% success rate on this, this is known technology. Instead, I average about one reboot per match. If this was a Blackberry Enterprise server I was running as my job, I'd likely have already been fired.
Our setup, our system, is all distinctly home brew. Hell, previously, I was running this on a Ubuntu box I learned to use in less than a week. The library lacks books on this specific subject, so I'm trying to learn from online sources. Apparently the big boys (like say, the local massive college sports department) have magic, $8000+ boxes that do their webcasting* by just being plugged in. That's called a "turn key" solution, is outside my price range, and anathema to my need to understand what's going on. If my fellow crew members want to take this show on the road, I've got to put this all together myself, and properly.
*(Or they're supposed to, I'm told by an informed source that sometimes they don't work and no one knows how/why)
0 Comments:
View My Stats
-->
