Just got off the phone with Rackspace concerning their planned outage on 18 July. After receiving the email announcing the planned outage:
’The Rackspace Cloud’s system engineering team will be performing upgrades to our storage infrastructure in our Dallas - Fort Worth (DFW) data center . . . During the maintenance period, you may be unable to access, upload or make changes to your website content. A "We are performing system maintenance" message will also appear on each of your websites.’
The problem? After viewing their sample outage page, I noticed it was putting out a ’200’ HTTP response header. I got a tech on the phone. After much shuffling and checking on their end, they confirmed that during the outage all affected cloud sites would display the outage message and a 200 HTTP response. No, the confirmed there was no plan and no recognized need to return 503s during this long outage. Yes, the technician confirmed, they realized this was not an optimal situation. If I or any of my sites were affected, I should discuss it with billing.
Ouch. Best practices and Google recommendations are below the clouds. A little less fanatical support, a little more attention to the fundamentals.
1 Comments
’The Rackspace Cloud’s system engineering team will be performing upgrades to our storage infrastructure in our Dallas - Fort Worth (DFW) data center . . . During the maintenance period, you may be unable to access, upload or make changes to your website content. A "We are performing system maintenance" message will also appear on each of your websites.’
The problem? After viewing their sample outage page, I noticed it was putting out a ’200’ HTTP response header. I got a tech on the phone. After much shuffling and checking on their end, they confirmed that during the outage all affected cloud sites would display the outage message and a 200 HTTP response. No, the confirmed there was no plan and no recognized need to return 503s during this long outage. Yes, the technician confirmed, they realized this was not an optimal situation. If I or any of my sites were affected, I should discuss it with billing.
Ouch. Best practices and Google recommendations are below the clouds. A little less fanatical support, a little more attention to the fundamentals.
1 Comments
Comments
A major supplier can’t (or won’t) implement 503.Idiots. WTF do they think the HTTP specs are *for*?