Scheduled maintenance for musclebeach.ladrupal.org at midnight on Sunday, May 21, 2011 (12:00am PDT / 7:00am GMT)

Events happening in the community are now at Drupal community events on www.drupal.org.
christefano's picture
Start: 
2011-05-21 00:00 - 02:00 America/Los_Angeles
Organizers: 
Event type: 
User group meeting

Hi everyone,

SoftLayer has a new datacenter in San Jose, CA, and I'm moving our VPS (which they call a CCI, or cloud computing instance) that hosts musclebeach.ladrupal.org to it from their Seattle datacenter. I've talked with SoftLayer's support team about this process and everything is planned out and ready to go.

When is this maintenance window?

This is scheduled for midnight on Sunday, May 21, 2011 (12:00am PDT / 7:00am GMT). It was originally scheduled at midnight on Saturday, but that conflicts with the Rapture. (Haha, that was to see if you're still reading. The real reason this is being postponed by 24 hours is because this scheduled maintenance was announced on the ladrupal-organizing mailing list and there wasn't any feedback. If you have any questions or concerns, please post them below.)

For everyone who wants to participate (seriously, though, this is like watching paint dry), I can be in #drupal-la during the scheduled maintenance. All I ask is that you review the existing ticket at https://manage.softlayer.com/Support/viewTicket/2993625

What sites and services will be affected by this?

The following services will be interrupted during this downtime:

  • Email to and from the ladrupal.org domain
  • MailMan mailing lists, including ladrupal-organizing
  • The ladrupal.org, intranet.ladrupal.org, ddcla.org and drupalcampla.com websites
  • Jabber, which hardly anyone uses

groups.drupal.org/la, #drupal-la and anything else living outside of musclebeach.ladrupal.org will not be affected.

Why move our server to California?

Latency of connections between the physical location of our server and that of the vast majority of our expected visitors will be reduced considerably. The average ICMP response from the Seattle datacenter is 40ms. In comparison, the average ICMP response from the San Jose datacenter is half that at just 20ms. This can be verified by pinging the following IP addresses:

   Seattle:
   208.43.138.186

   San Jose:
   50.23.105.18

This isn't an earth-shattering improvement but lowering network latency will noticeably improve the experience of the visitors and users of our sites and services that are in Southern California, our primary audience.

What is technically involved in this move?

This process moves the VPS image from one datacenter to another. Once the image has been cloned and deployed to the San Jose datacenter, the only required configuration "to go live" is to update the DNS to point the affected domains to the new IP addresses. More importantly, because this process involves cloning the entire server image, no data will be lost.

We can expect a maintenance window of 30 minutes to 2 hours.