Hello,
Please could I have some assistance. I have recently started to have issues when trying to match in Flukebook. I am receiving the “Attempting to fetch results” message, especially for matches using the MiewID v4 Matcher algorithm.
I have a fluke that has an identification status of “processing” and I see the results from the PIE v2 (Pose Invariant Embeddings) algorithm, but I have the error code for MiewID v4 Matcher. (Example below)
Then I have another fluke that has an identification status of complete but I receive the same error message for both algorithms.(E.g. - Flukebook )
Is this a situation where I need to give the program more time or is there a bug?
Many thanks, Ruth
Hi @ruth.wpa
Thanks for letting us know. I’m re-running these matches now and waiting for PIE results to finish. If they still have errors I’ll need to check with the ML team to keep troubleshooting.
encounter 1: https://www.flukebook.org/react/encounter?number=6b678c4a-f252-4865-a7f7-4084bf1a88cd
encounter 2: https://www.flukebook.org/react/encounter?number=60c2f639-f733-4c34-b094-b583050d096b
1 Like
I am also having a lot of these “attempting to fetch results”
Sometimes on one algorithm, often on both.
It happens on a lot of results of several recent bulk import.
I thought this was an internet speed issue.
What could I do to fix this?
I have kept all the links, but not the encounters related unfortunately.
Thank you!
What Wildbook are you working in? Flukebook
What is the entire URL out of the browser, exactly where the error occurred?
Flukebook
Can you describe what the issue is you’re experiencing?
On a lot of encounter results, the page isn’t loading and it showing “attempting to fetch results“
Hi @GwenBonaire
I re-ran ID on your encounter and it has results now: https://www.flukebook.org/react/encounter?number=a3c5deae-8c34-4f14-a2e9-0a709c471e1b
I also spot-checked a few of the encounters from the same bulk import and their match results appear to be loading correctly for PIE and MiewID. Let me know if re-running the match still doesn’t resolve the issue if you see other encounters from that import with missing matches.
@ruth.wpa We’re still working on why these matches aren’t fully returning even when they’re re-run. I hope to have an update shortly.
1 Like
Thank you!
When you are on the page I sent you, how to you go back to the encounter? Because nothing charges on the page, I don’t even have the link to the encounter:
I use one of our dev tools as a shortcut to find the encounter. In between the Flukebook URL and the task ID number, paste this in:
/obrowse.jsp?type=Task&id=
Your new URL will look like this:
https://www.flukebook.org/obrowse.jsp?type=Task&id=3513eefb-7a9c-47b4-a7ac-114b3bfbcab4
This brings you to the obrowse page for the specific matching job. It lists the algorithms run, which species it’s for, and which locations and IA classes it’s finding matches for. From here, click on the Media Asset Object link and then for the next page that loads, it will have a link to the encounter the media asset came from.
Hi Anastasia,
Thank you for your help. Today the results are coming through and I can complete the matching. I am not sure what changed, but everything seems to be working well now.
1 Like
Excellent! The devs found a bug that showed that the identification was completed, but Flukebook was not retrieving the results correctly. Thanks for your patience.
