I have build an API around Umbraco, where Umbraco serves as an easy way to update items and properties. I serve the data from my own API, this includes everything even images uploaded to Umbraco.
I'm gonna try to explain how everything is setup, the problem currently shows itself on the test server.
I have 3 environments; My local development machine, a development server and then a test server. I setup the database the first time on my development server and then moved it locally and then to the test server.
The problem shows itself like this; i have setup items, these items has properties on them and then an image to identify the item on the frontend. During the development I just uploaded some images and now when it's running on the test server I have uploaded the correct images. Everything works like a charm until the application pool recycles when the images suddenly rolls back to the old ones. Logging in to the backend the images are there, and if I republish everything the images that comes out from my API are correct again. I am suspecting that this includes all properties, but I can't really verify until the tests are complete.
This problem doesn't seem to occur on the development server and when debugging locally it doesn't happen either. So you could say I'm kinda stuck :)
I build the site with Teamcity on another server, so the only thing these servers once have had in common is the database.
I can recreate the problem by simply turning off the site, recycle to application pool and then restart the site. Anyone know where to start looking?
This sounds like the examine index is not up to date. Can you do a health check (on the developer tab in 7.5) or stop the site and delete the examine indexes under /app_data/temp/
Hi again! Sorry for this late response. That did not directly solve the problem. I was able to se a "warning" though about Umbraco not being able to write to web.config. And I can honestly not remember if we uploaded these images on the development server and then moved them to the test server, the same time that we moved the database. If this is the case can it be a permissions problem with Umbraco not being able to write to the cache local cache?
It turns out our VPN is down right now for some reason though, so I have to verify this tomorrow morning.
My application uses MSMQ to write to message queues on another server, so the API was running under Network Service. As it turns out Network Service had read permissions on everything, so the site worked. But it appears there were problems writing to the local cache.
I gave Network Service write permissions for now, might not be the safest but it will have to do for now.
Publish rollback when recyling app pool
Hi all!
I have build an API around Umbraco, where Umbraco serves as an easy way to update items and properties. I serve the data from my own API, this includes everything even images uploaded to Umbraco.
I'm gonna try to explain how everything is setup, the problem currently shows itself on the test server.
I have 3 environments; My local development machine, a development server and then a test server. I setup the database the first time on my development server and then moved it locally and then to the test server.
The problem shows itself like this; i have setup items, these items has properties on them and then an image to identify the item on the frontend. During the development I just uploaded some images and now when it's running on the test server I have uploaded the correct images. Everything works like a charm until the application pool recycles when the images suddenly rolls back to the old ones. Logging in to the backend the images are there, and if I republish everything the images that comes out from my API are correct again. I am suspecting that this includes all properties, but I can't really verify until the tests are complete.
This problem doesn't seem to occur on the development server and when debugging locally it doesn't happen either. So you could say I'm kinda stuck :)
I build the site with Teamcity on another server, so the only thing these servers once have had in common is the database.
I can recreate the problem by simply turning off the site, recycle to application pool and then restart the site. Anyone know where to start looking?
Thanks!
This sounds like the examine index is not up to date. Can you do a health check (on the developer tab in 7.5) or stop the site and delete the examine indexes under /app_data/temp/
Kind regards
Damiaan
Wow, it might just be that simple! I am seeing some errors on this screen, I will verify when the test server isn't in use and get back to you!
Good luck!
Hi again! Sorry for this late response. That did not directly solve the problem. I was able to se a "warning" though about Umbraco not being able to write to web.config. And I can honestly not remember if we uploaded these images on the development server and then moved them to the test server, the same time that we moved the database. If this is the case can it be a permissions problem with Umbraco not being able to write to the cache local cache?
It turns out our VPN is down right now for some reason though, so I have to verify this tomorrow morning.
Alright! That solved the problem.
My application uses MSMQ to write to message queues on another server, so the API was running under Network Service. As it turns out Network Service had read permissions on everything, so the site worked. But it appears there were problems writing to the local cache.
I gave Network Service write permissions for now, might not be the safest but it will have to do for now.
Thanks for you help!
Glad it is solved!
Damiaan
is working on a reply...