This Sunday, igniterealtime.org will move to a new server. The time window for moving the site is 12:00pm - 4:00pm (PST), but the expected downtime is fairly short. The server move will let us more rapidly respond to server issues and also to more easily roll out new services.
I hope Apache server’s issue (that i have to login 3 times) will be fixed along with that.
One of the improvements will be an upgrade to Apache 2.2, which I expect will fix that issue.
more rapidly respond to server issues
- Do you also monitor the server and forum actively or does one still need to send IM messages in case of problems?
more easily roll out new services
-
Do we also get an up2date SSL login and registration form?
-
Do we also get a Clearspace Widget which displays open_chat@conference.igniterealtime.org?
We do currently monitor igniterealtime.org, xmpp.igniterealtime.org and the Community pages, but we’ll be able to more rapidly respond now. Plus we’ll be able to more deeply monitor the server and the running services.
Oh also, we’ll be finally upgrading that old Fisheye instance to the latest release along with Crucible. And Subversion 1.5.
All moved! Let any of us Jivers know if you run into any problems.
Not all. I’m missing the Smack documentation
The server was down for hours today (4+ hours). I would say that does not bode well. What happened to this statement:
“We do currently monitor igniterealtime.org, xmpp.igniterealtime.org and the Community pages, but we’ll be able to more rapidly respond now. Plus we’ll be able to more deeply monitor the server and the running services.”
Adrian, docs are back.
Sorry Todd. We were notified of the downtime and were working on restoring service during that time. There were a number of small issues that resulted in the protracted downtime:
-
Apache misconfiguration, which is now fixed
-
Load balancer closing port 80 when it detected that the web server was not responding - I changed the health checks that our F5 does to determine if the server is alive or not, which should prevent it from closing the port prematurely.
-
We switched to using an HTTP connector in Tomcat versus AJP because connections weren’t being honored, which caused the downtime initially. We’re looking into why this happened to prevent it in the future.
-
Clearspace throwing NPEs with BlogServices, which caused 500 errors for the main web site. A rebuild of the search indexes and a restart of Community cleared this up.
Don’t let this get you down. We’re committed to maintaining a high service level and respond immediately when issues arise.If you ever notice anything awry feel free to send me an IM or an Email: chris.brentano@jivesoftware.com.