Goplan performance notes
At least a few of our users have been noticing a few performance issues when using Goplan - they may not affect everyone because of the way the application is clustered across our servers, but the bottom line is that some people have been affected with requests that take longer than usual, and even a few halts with a “Proxy Error”, which means that the server inside the cluster took longer than what’s acceptable to reply.
So here’s a status report and what we’re doing to solve this issue and make sure performance gets back to normality.
Current status
We’re actively looking at this problem, and things appear to be fixed. While I know it isn’t comfortable to read the words “appear to be” (instead of a better “things are fixed”), I also think that running circles around the issue isn’t helpful. Performance is back to normal, and we’ll monitoring the servers non-stop until we’re sure the dreaded “Proxy Error” won’t come back.
But why has it been slow? The increased recent load on the servers is a consequence of both the peak in traffic we’ve been experiencing and our new instant data backup and replication system - I don’t want to throw down any tech lingo on you guys, but it’s pretty cool and the bottom line is, it’s securing your data. We’re adding new machines to the cluster to bring the load back to normal and accommodate everyone better.
So, to recap: Things appear to be normal now, and we’re both making sure the problem did go away for good and that we add enough new machines to keep performance levels high. As always, we’re available on our support email should you need to voice any concerns - we are listening.
That’s good to know everything is getting back to normal.
Any clue when you guys start rolling out new features?
Comment by Maciek — August 14, 2007 @ 1:39 pm