The Coffee Desk will be going down for a major server move and some routine maintenance for around 45 minutes to an hour today starting around 4:00 – 4:30 Eastern Mountain Time. Our Twitter profile can be used to track the upgrade’s progress as it happens.

The move is largely a front-end server upgrade, moving from our old (yet faithful) dual Pentium III proxy server to a 64-bit, 4 GB RAM beast that will handle the next slashdotting better.

The backend servers will remain the same, as they have been for a very long time (and spanning several behind-the-scenes upgrades as well). The front-end server, however, is a much larger move and requires some actual downtime.

The move, for the curious, involves redoing the power and network structure used for the main server HTTP proxy front end, as well as configuring NFS, TFTP booting and IP switches between the old and new server.

This new server I keep speaking of will be mirroring the configuration of the old(e) server, only with a tremendous (and overdue) processing power upgrade. In addition, the new server is considered to be “greener” than the other beast (not that we care that much), and also quieter in operation (not that you care that much).

So, we apologize for the downtime disrupting the RSS updates for our dedicated readers, and for our new readers: this is the only downtime we’ve had since the start of the site other than some slashdotting/bandwidth issues.

We aren’t like these amateur kiddie sites that go down because mommy forgot to pay the DSL bill.

So we’ll be back up in no time, and keep up with our Twitter for upgrade updates. We’ll be back in a bit, and with more articles too.

Post-Upgrade Status

Well, as some of you noticed, the move didn’t go so smooth although everything is back online now.

Everything was moved over within the 45-minute time frame as planned as far as HTTPD, NFS, and various hardware modifications went, but the database move took much longer.

Part of it was my fault. I didn’t backup the database properly to the temporary NFS storage server, and upon realizing this it became necessary to bring the old server back online again to perform a proper backup of the various databases.

Then, a MySQL permissions issue that occurred as a result of importing the database fought me all the rest of the way. I finally solved it, after scouring through various mailing lists and USENET posts, but the effort involved in fixing it was almost unexcusable.

But we’re back online now, and new articles are coming shortly. Our beautiful year of uptime took a hit as a result of the upgrade, but it was very necessary for continuing our operations here – trust me.