Collector offline download hangs at 50%, never finishes

4990
7
Jump to solution
01-30-2015 02:35 PM
RoyceSimpson
Regular Contributor

I'm trying to get Collector work offline with a feature service hosted through our 10.3 ArcGIS Server and 10.3 Oracle SDE database.  I've configured and followed the "prepping data for offline" docs.  I've enabled attachments and have users setup in AGS for the "private" feature service.  Editing the data within the feature service works well in every instance... In ArcGIS Online, works fine, in ArcMap works fine, Collector works fine.  But when I try to tap the "download" button in the Collector app on my Android phone, it always gets to about 50% just just hangs indefinitely.  I've inspected my AGS Jobs folder and can see submitted jobs for the replica.  I have no idea how to properly troubleshoot this.

Thanks much,

-Royce

Tags (1)
0 Kudos
1 Solution

Accepted Solutions
RoyceSimpson
Regular Contributor

Thanks Russ,

Ok, so the null field problem was because I had my mxd setup with a connection to the geodb via a user that wasn't the owner of the data.  As soon as I remapped the connection to the feature service layers back to the owner of the feature classes, the null field error went away.  I did in fact have no fields that were non-nullable so why that error even came up was a mystery.

Everything seems to be working fine now.  I can download and edit offline and syn properly.  I even tried registering the data as versioned and was able to have multiple users on different versions do edits and sync.  Very cool.

Thanks for your help.

-r

View solution in original post

7 Replies
RussellRoberts1
Esri Frequent Contributor

Enable debug logging on your server and see if any errors are showing up in logs. You can also try to create a replica through the REST endpoint and see if the a replica is being generated successfully from the service directly.

Russ

RoyceSimpson
Regular Contributor

Hi Russell,

I've done the debug thing and don't see any errors.  There are tons of messages but nothing I can latch onto.  I don't see anything in the feature service's output folder so I'm guessing nothing is actually coming out of the replica creation process.  I just don't know how to move forward on figuring out why.

I've looked at the "create replica" on the rest endpoint but have thus far been too lazy to go through all the json hoops to try to do that.  Guess it's time to do so.

-r

0 Kudos
RoyceSimpson
Regular Contributor

Ok, I submitted an async "create replica" command via the rest api... I obtained a "check status" URL for the job but when I go to the URL, I get a blank page, literally, there is no HTML source on the page. 

0 Kudos
RoyceSimpson
Regular Contributor

I tried again, this time as a "synchronous" job.  Server returned a URL with a sqlite .geodatabase file... seems to have worked. 

So, the question is, why is async method not working?

0 Kudos
RoyceSimpson
Regular Contributor

Unfortunately, now, when I try to download in Collector I get an error, "...failed to download with the message:  Unable to complete operation.  Bad syntax in request".

However, If I turn "async" back on for the sync service and check the "View results with a map service" option.  It works, downloads.

But... when I create new features and "sync" from Collector, it does it without errors but no features show up in the main geodatabase.

UPDATE:  getting this error, "

SEVEREFeb 2, 2015, 11:17:37 AMFailed to transfer inserts for CollectorPoints. The field is not nullable.System/SyncTools.GPServer"

Not sure how to fix that one.  What field is not nullable?

0 Kudos
RussellRoberts1
Esri Frequent Contributor

This should be one of the fields on the feature services does not allow null values and somehow it is sending a null value up to the service when you are syncing.

Also – on one of the posts they try the async call with the REST page – they need to for a format=pjson. The HTML page may be blank until there is a result.

Russ

RoyceSimpson
Regular Contributor

Thanks Russ,

Ok, so the null field problem was because I had my mxd setup with a connection to the geodb via a user that wasn't the owner of the data.  As soon as I remapped the connection to the feature service layers back to the owner of the feature classes, the null field error went away.  I did in fact have no fields that were non-nullable so why that error even came up was a mystery.

Everything seems to be working fine now.  I can download and edit offline and syn properly.  I even tried registering the data as versioned and was able to have multiple users on different versions do edits and sync.  Very cool.

Thanks for your help.

-r