mcneel/compute.rhino3d.appserver

A node.js server for solving Grasshopper definitions on Rhino Compute

compute.rhino3d.appserver's Issues

Handle cases when caching is not desired

• In situations where compute results are errors or empty data
• In situations where one wants to recalculate
• In situations where one doesn't want the result cached

Link to "Developing with Rhino Compute" in documentation doesn't work anymore

Hey,

the link to the guide is broken.

Update readme to include workshop video

Workshop video now available at https://vimeo.com/442079095
Probably want to also include slides from workshop at
https://docs.google.com/presentation/d/1nCbd87iA_D2ZCwoSirOYK3har6XUJHDUEIkt635btUU/edit?usp=sharing

Run npm i (or npm ci?) as part of vscode launch configuration

Simplifies set up! Clone, F5 and go!

srf_kman clustering. Different output from what I see in Rhino

Hi there.

I was playing a little bit with Rhino.Compute and I would raise an issue regarding the srf_keman clustering output.
Does it work as expected in your computer/server?

Support GET request for solving

Add /solve/{definition} with support for query string inputs for solving

This would let us implement caching headers for browsers as we could return the hash of a definition in an etag
See https://devcenter.heroku.com/articles/increasing-application-performance-with-http-cache-headers

investigate etag header values

https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/ETag

Catch server errors

More gracefully catch 500 Internal server errors from Compute Server. Throwing an error results in an UnhandledPromiseRejectionWarning.

add job queue

Handle cases for long running solutions.
Are there job queues strategies that do not add another dependency like Redis?

A question about exporting step file

Dear all：
I have a question about exporting step file.
I called the GH file by computing service（https://developer.rhino3d.com/guides/#compute）,
but I could not export the STEP format using the LunchBox save Object plugin .
While I can export STEP format using the same plugin by openning grasshopper software.
could you tell me the right way to call GH file by computing service?
I look forward to your reply as soon as possible.
Thank you very much.

Use memcached on Heroku

Provide an option in our appserver caching to use memcached instead of our current in-memory cache. We would need to include something like the hash of a definition contents in the key so we don't get old cache results from a previous run of the appserver and an old GH definition.

Seems like we could check for a memcached environment variable on Heroku and if it exists use that instead of the in-memory cache.

Test appserver locally against a remote compute server

I have appserver running fine for production. We have a compute server in a virtual machine and we have appserver on a separate web machine.

For locally testing during development I am trying to run appserver locally, while still connecting to our remote compute server so that I do not have to install Rhino on my local machine. So I am doing:

export COMPUTE_URL=some external server
export RHINO_API_TOKEN=…
export RHINO_COMPUTE_KEY=…
npm start

But, this does not seem to work, I keep getting 500 errors from the compute server. Should I expect this type of setup to work? Does the compute server require some sort of publicly accessible callback URL at appserver?

Thanks for any thoughts. Andrew.

Implement cache warming scripts

Allow for creating a script that can be run which can asks the appserver to solve many different scenarios in order to build up a large set of cached results. The idea is that if we had a definition that took a minute to solve for each iteration we could create a script that ran a large series of parameters and we could have 1,000+ cached solutions after a day of hitting compute. Cached results will take milliseconds to return instead of a minute each. Load balancing out to multiple compute instances would increase this number per day

Hosting FEA tool on Compute

Hello.

I am trying to explore the use of a web app to solve some simple FEA Grasshopper model but I am having some issue.
The tool required to run an external application (I do it with subprocess) from which we can read an output file.
Is it possible?
I am getting this error message but I am not able to get any tip.