Update on Performance Issues
Last Thursday, we posted on this blog explaining any recent service snags in the previous few days.
Since then, we’ve done a number of things to pin down and address the issue. It turns out that the issue was due to our Nginx load balancing server hitting the bandwidth cap as configured. Let me explain briefly:
Our load balancing server is like a traffic cop directing traffic to the right roads. In this example, the roads are the web servers handling Disqus usage. The issue I described earlier is as if the traffic cop just passing out (he has an abnormal heart rhythm). With the traffic cop incapacitated, the traffic doesn’t know where to go even if there is plenty of room on the roads. You’ll have to stretch your imagination to accept such dumb drivers, but I hope that paints a decent picture.
In addition to addressing this, we are constantly adding capacity and rewriting parts of the service to better handle the growth.
On the subject of improving Reactions, which was mentioned in the previous blog post, you will be able to see the improved results of our changes within the next couple days.
Daniel on June 1st 2009 in disqus
-
maxomen
-
MeowBowWow
-
totallyace
-
beats
-
chiropractic coaches
-
Traveller
-
tisdevendra
-
muckdog
-
chat
-
muckdog
-
muckdog
-
muckdog
-
monkeyfrog
-
monkeyfrog
-
MovinFwd
-
muckdog
-
monkeyfrog
-
MovinFwd
-
Skip
-
shana
-
afroed
-
Lawrence_Chiu
-
Ed Maximus
-
Skip
-
monkeyfrog
-
afroed
-
Lawrence_Chiu
-
Lawrence_Chiu
-
Giannii
-
shana
-
shana
-
giannii
-
shana
-
Guest
-
emilcohen
-
Giannii
-
lostremote
-
lostremote
-
giannii
-
Brian McManus