Did you also update to the latest version of the AWS A3 Picker?
It looks like the CDN Domain is missing in the Now-snippet. Could you try and re-save the settings on the datatype and re-select + republish one of the videos to see if its updated correct?
I believe the issue only occurs for some files, which results in missing cdn name. I have located and fixed the issue and will upload an updated package very soon.
I've been getting the same issue, but after downloading the new version it's fixed that problem. Though I'm getting a new issue though.
When I upload a file using the S3 provider it will always upload to the root folder, even if I select a subfolder under the root.
I also noticed you didn't package up the AWSSDK.dll in the new install, since you're using an older version of the AWSSDK(current release is 1.3.8, the one used for the S3 provider is 1.3.1) thought I'd let you know.
Is there a place to get the source for this project? as I could write a patch for it if you're a bit busy.
Thanks for the update. That fixed the issue for me regarding the missing cdn name. However, I am also having the problem where all files are uploaded to the root of the bucket even when you specify a folder. It is not really a big deal, but I was going to make directory for images and videos to keep them seperate. Can you check and see if it is doing the same for you.
As far as updaing existing records to include the cdn, do you know of a quick way that I could "update" my exising documents that are missing the cdn. I had put a work around in my templates to add the cdn if it was missing as a short term fix. Now that that bug is gone, I have about 100 documents that clients uploaded and picked videos that don't have the cdn. Is there something I can do in the DB or via the API to quickly get all the documents and update them.
AWS S3 Picker is now updated. Version is the same 1.3 as before.
I dropped merging the dlls, and found an issue with the folder structure when no new foldername was entered. This has now been fixed, so the uploaded files doesn't end up in the root (unless uploaded to the root of course).
Let me know if you encounter anything else.
@Justin in order to update existing nodes I believe it would be sufficient to Republish the node (you might have to Save and Publish, but dont think its necessary). So if you create a script that loops through your nodes and publishes nodes with the AWS S3 Picker datatype on it you should be able to fix your existing properties. Let me know if you need help with that.
Thanks a lot for the update, this has fixed the issue with the folders not being created/media not being saved to them correctly.
Though I'm getting one particular error when viewing/selecting a value, it seems that the picker is formatting the URL to the image incorrectly when using the choose tab and using the S3 provider.
It formats the image path as http://{bucketname}/{folderpath}/.s3.amazonaws.com/{folderpath}/{imagename}
The folder path is written into the URL twice, an example of how it's being formatted is.
Where Test is the folder inside my bucket, and thumbnail.jpg is the asset name. this is how the path is saved to the database which means at the moment all the image paths are being saved incorrectly and I have to use a helper function to correct them.
Universal Media Picker 1.5 with Amazon S3 missing data
Our Picker has stoped loading the data correctly
Before :-
Now :-
all the settings in the data type have not been changed, has any one experienced this before
or know of any updates that could have caused this problem
Thanks
Hi Clive,
Did you also update to the latest version of the AWS A3 Picker?
It looks like the CDN Domain is missing in the Now-snippet. Could you try and re-save the settings on the datatype and re-select + republish one of the videos to see if its updated correct?
- Morten
Hi Morten
I had to launch the site with a get around,
but I'll give that a go, as we are using this combination on a number of our sites
thanks
Hi Morten,
I have just confirmed this also. On our development site and production site, the cname is not coming through using UMP 1.5 and S3 provider 1.2.
I have resaved the data type and altered the cname and altered back again.
Any thoughts?
Regards,
Jeavon
Hi Jeavon,
I sent you an email with updated assemblies.
Let me know if it works.
- Morten
I'd like to know if that works too. I think that my site was having the same issue after I upgraded recently to the latest UMP + S3 provider.
Hi Justin,
It does work! I'm sure Morten will release the updated v1.2.1 S3 provider shortly.
Cheers,
Jeavon
Hi Justin,
I believe the issue only occurs for some files, which results in missing cdn name. I have located and fixed the issue and will upload an updated package very soon.
- Morten
Great! Thanks for the quick action! You rock.
Hi Justin and Clive,
I have uploaded version 1.3 of the S3 Provider, which fixes the CNAME issue.
Sorry about the wait.
- Morten
Hi Morten,
I've been getting the same issue, but after downloading the new version it's fixed that problem. Though I'm getting a new issue though.
When I upload a file using the S3 provider it will always upload to the root folder, even if I select a subfolder under the root.
I also noticed you didn't package up the AWSSDK.dll in the new install, since you're using an older version of the AWSSDK(current release is 1.3.8, the one used for the S3 provider is 1.3.1) thought I'd let you know.
Is there a place to get the source for this project? as I could write a patch for it if you're a bit busy.
Cheers,
Ryan
Hi Morten,
Thanks for the update. That fixed the issue for me regarding the missing cdn name. However, I am also having the problem where all files are uploaded to the root of the bucket even when you specify a folder. It is not really a big deal, but I was going to make directory for images and videos to keep them seperate. Can you check and see if it is doing the same for you.
As far as updaing existing records to include the cdn, do you know of a quick way that I could "update" my exising documents that are missing the cdn. I had put a work around in my templates to add the cdn if it was missing as a short term fix. Now that that bug is gone, I have about 100 documents that clients uploaded and picked videos that don't have the cdn. Is there something I can do in the DB or via the API to quickly get all the documents and update them.
Thanks,
Justin
Hi Guys,
Sorry about the updated package not working. It seems the merge of the AWS dll and my own dll has gone wrong.
I'm working on a fix now.
I will post back to this thread when its updated.
- Morten
AWS S3 Picker is now updated. Version is the same 1.3 as before.
I dropped merging the dlls, and found an issue with the folder structure when no new foldername was entered. This has now been fixed, so the uploaded files doesn't end up in the root (unless uploaded to the root of course).
Let me know if you encounter anything else.
@Justin in order to update existing nodes I believe it would be sufficient to Republish the node (you might have to Save and Publish, but dont think its necessary). So if you create a script that loops through your nodes and publishes nodes with the AWS S3 Picker datatype on it you should be able to fix your existing properties.
Let me know if you need help with that.
- Morten
Hi Morten,
Thanks a lot for the update, this has fixed the issue with the folders not being created/media not being saved to them correctly.
Though I'm getting one particular error when viewing/selecting a value, it seems that the picker is formatting the URL to the image incorrectly when using the choose tab and using the S3 provider.
It formats the image path as http://{bucketname}/{folderpath}/.s3.amazonaws.com/{folderpath}/{imagename}
The folder path is written into the URL twice, an example of how it's being formatted is.
http://testbucket.media/Test/.s3.amazonaws.com/Test/thumbnail.jpg
Where Test is the folder inside my bucket, and thumbnail.jpg is the asset name. this is how the path is saved to the database which means at the moment all the image paths are being saved incorrectly and I have to use a helper function to correct them.
- Ryan
Hi Ryan,
Did you try this both with and without a custom domain?
- Morten
is working on a reply...