Site Problems on Sunday

[danielb]

During Sunday about 25% of our visitors experienced problems when visiting the site, as one of the machines in our cluster was having problems.

The reason for this was the recent introduction of a High Availability Component that allows us to quickly increase the number of machines in the cluster in response to high loads... When the load increased on Sunday, the system indeed added a machine but the machine wasn't working properly. Accordingly, visitors routed to the new machine experienced problems with accessing the site.

Like with anything new, the High Availability system may still have a few problems which we will iron out in the next couple of weeks. It's a change for the better even if in the short term it may create some problems.

Thanks for your understanding.

 

[Shane]

Just FYI Daniel. Sunday is probably the worst day to introduce new or upgrade anything. Anyway you could pick another day?

 

[Chaz]

Just FYI Daniel. Sunday is probably the worst day to introduce new or upgrade anything. Anyway you could pick another day?

I think he is saying that the problem was encountered on Sunday because of the traffic and a server failure, not that the changes/updates were made on Sunday.

 

[AZZenny]

True, but Sunday is the start of the work week there, so it's worth mentioning just for future reference.

 

[Linderbee]

yeah, I couldn't get on Sunday, but it was LATE. I thought it odd for the site to be down at that hour, but no big deal. Thanks for the update.