WeeklyUpdates/2012-04-09: Difference between revisions

Jump to navigation Jump to search
No edit summary
Line 131: Line 131:
* This week we turned up hundreds of new systems (more than I can collectively count amongst all teams in IT).  For example, 128 nodes with new Sandy Bridge Xeon procs have been deployed to the SCL3 data center, and 64 to the PHX1 data center.
* This week we turned up hundreds of new systems (more than I can collectively count amongst all teams in IT).  For example, 128 nodes with new Sandy Bridge Xeon procs have been deployed to the SCL3 data center, and 64 to the PHX1 data center.
* We replaced the NIC cards in our PHX1 load balancers this week with new vendor-recommended cards to overcome problems that have plagued us for months. So far, we have not had any lb-related problems in PHX1.  However, part of this success is due to moving the load of Addons version checking and about:home snippets to a new set of Netscaler load balancers which handle the heavy-loaded sites better but don't provide some of the features we enjoy out of our Stingray load balancers.
* We replaced the NIC cards in our PHX1 load balancers this week with new vendor-recommended cards to overcome problems that have plagued us for months. So far, we have not had any lb-related problems in PHX1.  However, part of this success is due to moving the load of Addons version checking and about:home snippets to a new set of Netscaler load balancers which handle the heavy-loaded sites better but don't provide some of the features we enjoy out of our Stingray load balancers.
* A lot of work was wrapped up in prepping a new environment for hg.mozilla.org in our new data center.  We should be announcing downtime soon to make the switch this week.
* ''hg.mozilla.org''
* Speaking of hg, this is a public service announcement reminding you not to ctrl-c a push to try once it has begun. Due to the added pushlog in commit hooks, it causes hg to break and requires manual intervention to fix.
** A lot of work was wrapped up in prepping a new environment for hg.mozilla.org in our new data center.   
*** We should be announcing downtime soon to make the switch this week.
** <i>Please</i> do not abort a push to try (with ctrl-c) once it has begun. Due to the added pushlog in commit hooks, it causes hg to break and requires manual intervention to fix.
 
* This week we stood up a cluster (admin, dev, stage, and prod) along with global caching for the mozilla-hosted Firefox 3.6 start page: https://wiki.mozilla.org/Websites/Firefox_3.6_Start_Page
* This week we stood up a cluster (admin, dev, stage, and prod) along with global caching for the mozilla-hosted Firefox 3.6 start page: https://wiki.mozilla.org/Websites/Firefox_3.6_Start_Page
* Data center migration in China for Mozilla Online continued this week, with Shyam on site finishing the load balancer and vm hosting, and working on moving key sites.  Mail should take another week and we will start wrapping up and heading home.
* Data center migration in China for Mozilla Online continued this week, with Shyam on site finishing the load balancer and vm hosting, and working on moving key sites.  Mail should take another week and we will start wrapping up and heading home.
Confirmed users, Bureaucrats and Sysops emeriti
1,081

edits

Navigation menu