All feed generation and delivery has been moved to the new server. You do not need to make any adjustments to your frames, they will automatically switch to the new server over the next few hours.
The web site is still running on the old server so will continue to experience slow response until I finish moving that as well.
This should resolve several issues with sporadically missing photos, invalid RSS feeds and general wonkiness. Please let me know if your frame is still unhappy.
FrameAlbum is a victim of it’s own popularity. The tiny server I have running the service isn’t up to the task now that there are several hundred people on the service. This is the reason that some of your frames are reporting ‘invalid RSS feed’ or ‘cannot contact service’ or similar messages.
My bad, sorry about that.
The server I’m using is based on Amazon’s EC2 service. This matters not to most of you but will be interesting to the .5% of you that are are geeks like me. The ‘tiny’ instance type is a small fraction of a ‘real machine’ and when someone else on the same physical machine as my instance runs something heavy it takes away from me. Well, it seems that someone is almost always running something heavy on the same machine as the FrameAlbum service. As a result, the service is a bit wobbly. I’ve tried stopping/starting the server a few times to get on a different physical machine but the problem persists.
I’m moving the service to another machine that is not shared. This will cause the service to be a bit bumpy for the remainder of the day but will result in a much more stable and responsive service.
Thanks for your support and patience!
I’m running the FrameAlbum service on a very modest server. I limited all the channels to a maximum of 100 photos to reduce the load on the server.
I’ve moved parts of the service to a new, more capable server. I’ve increased the limit to 200 photos to see how this new machine performs. Assuming it performs well I’ll bump it up again in the near future.
Enjoy!
I’m working on a database issue that is causing the system to generate invalid RSS feeds. I’ll post an update when it has been resolved.
Update: All better now.
Those of you using the FrameAlbum DNS server to redirect your frame may have encountered a bit of trouble today. In this context ‘a bit of trouble’ means total failure of your frame to retrieve the feed from the FrameAlbum service. ;-( It may have exhitibed itself as a ‘DNS ERR; RSS invalid’ error, or simply ‘unable to contact service’ message.
I am in the process of building greater redundancy into the system and inadvertently mis-typed a password used by the DNS server to access the database server. We in the business refer to this as a ‘fat finger’ error. 😉
The issue has been resolved and the DNS server is back to normal operation.
Sorry about that.
I pushed some new code to the FrameAlbum website today. This is a small update to improve the messaging on the status of a channel. It will now inform you if the system was unable to find a matching username on Flickr/PicasaWeb. It will also tell you how many photos were added to a channel during the last update cycle.
As always, feedback is appreciated.