[BUG] node/update is very slow/ non-terminating for very large skeletons

[clbarnes]

• CATMAID version: 2021.12.21.dev0+unknown
• CATMAID instance: https://neurophyla.mrc-lmb.cam.ac.uk/catmaid/drosophila/l1/igor
• Browser: Firefox, Chromium
• Operating system: Manjaro linux

Short description

Among other things, updating a node's location in a very large skeleton results in a call to node/update which does not terminate in reasonable time (minutes at least).

To reproduce

• Select a large skeleton
• Move one of the nodes

In this case, neuron "Tissue_whole-larva", 225 603 nodes. This is, notably, smaller that some of the skeletons in FAFB (see #2000 ).

More information

The browser detects the node/update request going out, so it doesn't seem to be the frontend which is holding things up. However, the change is not persisted to the database (when you reload or open a new instance), so it's not getting that far.

In firefox, the hang happens before the dev tools can even gather information about the request: in chromium, the request details are gathered.

[clbarnes]

This particular skeleton is just an outline of the body, presumably for mesh generation. This can equally be done with several smaller skeletons, so in this instance just splitting this would hopefully allow the team the continue working with the data. Not sure if there's a good way to do this on the server directly.

[clbarnes]

Minor update: the request "resolves" after 10 minutes with a 405, presumably because of a 600s timeout which I believe is the default in some layers of nginx.

[clbarnes]

The motivating problem seems to have resolved itself, suggesting that something external was being a problem (LMB network, maybe?).