Twitter is really working on it
I have been harsh with Twitter because of all the outages they experienced over the last week, but it is also worthwhile to point out the good things they have done.
Shifting hosts, though abrupt, clearly shows that they are trying to be pro-active and that they are more than aware of the dangers of their current system. Given last night’s outage, it remains to be seen if that was a worthwhile move, but we also have to acknowledge that these are complex systems and that recovery doesn’t happen from one day to another.
More importantly, if you have a look at their jobs page, they are explicitly looking for engineers that will focus on availability:
Operations Engineers
- Focus on consistent availability and reliability (self and systems).
- Excellent and influential communication skills (team, vendors, community).
- Experience with: Linux, OS X, Solaris, HTTPD (Apache, lighttpd, nginx), mongrel, tomcat, MySQL, pgsql, storage systems (NFS, NAS, SAN), load balancing systems (hardware, software, dns), automated system tools (cfengine).
- Experience maintaining: large number of servers, multi-tiered environments, monitoring and notification tools.
- Experience executing: well planned and tested solutions to complex problems, goal oriented workdays with multiple context changes and interruption.
- Must have: excellent triage skills, mild manner, rockstar inside (ready to rise to any occasion), strong interest in Twitter.
Whoever gets those jobs might be in for a hard time at first, but given the ever growing Twitter eco-system, uptime must be Twitter’s #1 priority right now and its good to see that they are working on it.
Cartoon courtesy of GapingVoid
2 Comments