Spot mapping whale sharks in Sharbook, two encounters in a row now I have received the message “Unable to fetch data: 200 - success” once I get to the modified groth results page after the scantask grid. Does anyone know what this means? It does not show any possible matches for the shark, just empty tables and this error message.
If it makes any difference - the encounters were from 2012. I am working on getting through our organization’s backlog of un-spot mapped encounters.
I tried an encounter from 2015 and had the same result.
Also had the same result for a 2017 encounter.
Now the same result for a 2021 encounter. Is the server down or something? This has never happened before, and has now happened with every encounter I’ve tried to spot map today.
Sharkbook is undergoing server maintenance. Keep an eye on this post for updates on when it’s resolved: MAINTENANCE:OUTAGE: Machine Learning Server Behind Multiple Wildbooks
Hello @Anastasia, thank you for your reply. Is there any estimate on how long the server will be down? Just trying to plan out my schedule around when Sharkbook will be back up and running. Thanks so much!
Ideally it wouldn’t have been down at all! This wasn’t planned and we don’t currently have a timeline to completion. We hope to have an update on the main post soon.
The Sharkbook server is back up and running. Can you let me know if you’re still running into issues viewing your scantask results?
I just tried spot mapping a couple, and the newly spot mapped ones seem to be working, but the ones I had tried to spot map when the server was down (I tried several throughout the course of trying to figure out what was going on) still come back with the same error. Even when I re-spot map them. One additional question - previously, the Suggested New ID would be in the MX-#### format. Now it is just suggesting three digit numbers, without the MX prefix, do you know the reason for this?
For further explanation - On the ones I spot mapped unsuccessfully before I knew the server was down, running the scantask again brought me to the same error. So then I used the remove spots function, and re-spot mapped the shark. Once I ran the scantask on the re-spot mapped shark, it never appeared in the “My Pending Scantasks” row, and just immediately went into the middle of the shark grid. When I viewed it from there, it was once again the same error. Is there any way I can spot map these sharks without deleting them and re-uploading the encounter?
Can you share the URL for some of the stuck scantasks so I can research this further?
Here is one of them
Thank you! I hope to have an update for you soon.
I’m sorry I didn’t address this one sooner. If you assigned new IDs outside of your project, it won’t suggest IDs with your project prefix. If this happened within a project, can you let me know which project you were working in when this happened?
We’re unsure how to get this particular Encounter cooperating again. I tried resetting your spot map and matching from a new one, but the same issue occurs.
Occasionally even when the original cause of an error is fixed (in this case, the server outage), jobs that were caught up in it don’t manage to reset correctly even with manual intervention.
Wildbooks recognize pixel-for-pixel when it’s seen an image before, so I’d like you to try something: If you delete this Encounter and then change the image slightly (such as resize it by 1 pixel) and re-upload it, it should treat it as a never-before seen image and hopefully stop associating it with whatever errors it associated it with previously.
OK, thank you so much for your help. It did not work for me either when I would delete the spots and run the scantask again. But now that I know what to do, I can delete and re-upload these encounters with the image changed slightly as I go. I will let you know if that approach doesn’t work.