…and maybe last classes, at least for a while. (And just when this was getting really fun!) Yesterday, I told my new journalism students that the class was paperless, that we were going to use just digital paper, and that all of their work for the quarter would be compiled in their Web logs. (One student actually claimed to a) know what a Web log was, and b) have her own up and running, though she couldn’t remember where it was…sigh.) I started them through the setup process, creating a site, filling in names and passwords and e-mail addresses, and I was feeling good now with, what, 75 or so Manila Web logs under my belt, and I said “Ok, just click submit and create your site!” and they all clicked and sat there giddy with anticipation watching the little IE flag waving in the corner waiting for their carefully chosen templates to appear, waiting, waiting, waiting…
Unbeknownst to me our little Frontier server decided not to even try to process the creation of 22 sites at the exact same moment and instead opted to stop all forms of digital communication to anywhere in protest. I’m pacing behind my kids, looking at screens, and from the deepest recesses of my brain comes floating up this vague, uncomfortable deja-vu type feeling like I’ve been here and done that, the energy of the moment just draining out through my toes. I call Ed the tech guy and he checks the server. “Didn’t this happen last fall?” he says…
Good news is about 10 of my kids actually created their sites from home last night. Today we get back on the train, slowly, a few at a time, to start another most excellent Web log adventure…stay tuned.
Sarah Lohnes says
This happens to us, to, when creating new sites all at once. Actually, it happens when we’re trying to create 6 or 7 sites at once! Anyway, the database tends to get bogged down. I was just made aware of a fix for database sluggishness, that a colleague found somewhere in the Manila docs:
As a Frontier object database file, Frontier.root, gets older, freed disk
blocks are added to an internal list called the “avail list.” As this list
gets long, disk-based operations can slow down significantly.
You can periodically freshen your database with the Save A Copy command in the
File Menu on a regular basis or when you notice a guest database is quite
large (a few hundred MB).
First, shut down the Web server by choosing Web->Web Server->Stop while you
compact the root files.
Choose the command, save to Frontier.fresh.root. Take a short break. When it’s
finished, quit Frontier, move Frontier.root to a safe place, and rename
Frontier.fresh.root to Frontier.root. Relaunch Frontier. Things should run
You can turn the Web server back on and trash the original Frontier.root. The
new one contains all the information in the old one.
If you want to get an idea of the condition of your object database, enter
window.dbStats () into the Quick Script window. A report appears in a separate
window. The third line in the stats window tells you how many nodes are on the
avail list. The fewer the better.
Save A Copy will also compact the file, squeezing out all the unused space
between disk blocks. The resulting file will be smaller than the original file.