Hey everyone… My server at http://nuxx.net is down, and no, I don’t know why. I think Waveform is moving locations, but they didn’t notify me, so I’m just not sure what’s up. I have a support ticket open about it, but they haven’t replied yet. So, we’ll see… I hope I don’t have to move providers.
So, the Best of Bootie 2007 comp I hosted a mirror of and linked to yesterday? I had to remote it. My LJ post was linked to from elsewhere (sorry, can’t tell from where) and that was causing just a little too much bandwidth to be used.
If you still want this and know how to contact me directly, do so. Otherwise, please just grab it from the Bootie site.
Er, my Xbox 360 Chatpad Disassembly photos (photo gallery retired) have made it to Kotaku and Xbox-Scene.
I’ll be upgrading some things on nuxx.net over the next few hours, so anything hosted there (including email, as SpamAssassin needs an upgrade) may be temporarily unavailable.
I’ll update this post once things are done.
Know what’s crap? LJ’s newfound ability to embed any media, including auto-playing Flash stuff.
Now when opening my friends page there is a very good possibility that there will be some crap music / skit / something someone thinks is funny just automatically playing away.
Upon further investigation it seems that there is a way to have embedded things shown as placeholders. I’m enabling this now.
So, earlier today I submitted my Lightsticks In The Toilet LJ Post and photo album (photo gallery retired) to the Make Blog. I had also submitted it to Digg sometime last night.
Well, it has been posted on the Make Blog, and the Digg link has really taken off. With a total of 5 fastcgi processes for myself (nuxx.net, et al) the box is sitting at a load of ~5 to ~10, and is handing out loads of data. Feel free to watch the bandwidth consumption here.
This is interesting. I’ve only had to deal with tremendous bandwidth load before (eg: hosting the SG videos, or original seed for Katrina photos), not something that actually loads the box.
I just moved the photos to nine per page (3×3) to cut down on the impact of each page load. We’ll see what happens, I guess…
I’ll be upgrading PHP on rowla.nuxx.net a bit later today, so the hosted website(s) will go down for a while.
To everyone who uses rowla.nuxx.net to some degree: This machine is going to be moved at some point within the next few weeks, hopefully within the next few days. The colocation provider I use is getting their new building / data center going, and I’m going to move in there sooner than later. Because of this I will have to switch to a new IP address and have the server down for a few hours while it is driven to the new building, but after that things should be fine. It should just come back up once turned on.
Email deliveries attempted while the server is off should be queued by the last-hop mail server and delivered after things come back online.
I’ll post more info (and more specific dates / times as I have them) here, but for now just know that it’ll be happening.
Ah, I’ve now done a write-up on nuxx.net. So, feel free to give it a read if you are curious what sort of things are running behind the scenes on the box which hosts my (and others) sites to… well… host things.
For some reason (probably because of the size of my gallery and the number of sites hosted on the box) I’ve been having problems with Google being able to pull an XML sitemap for the gallery. Google was apparantly timing out while attempting to download it. I’m not sure if some change was made on their end recently regarding the length of time which could be spend during sitemap downloads or what, but I think I found a solution:
#!/bin/sh
newfile=~/www/nuxx.net/gallery/sitemap.xml.new
oldfile=~/www/nuxx.net/gallery/sitemap.xml/usr/local/bin/curl -Sso $newfile http://nuxx.net/gallery/main.php?g2_view=sitemap.Sitemap
if [ -d ~/www/nuxx.net/gallery/sitemap.xml.new ]; then
mv $newfile $oldfile
fi
Yes, I know it’s a hack, but running it every six hours should work around the problem.
UPDATE: Yep, I just realized this won’t work. The test argument needs to be changed to -f (or -e) because it’s not a directory. Doh!