Archived from the Transloadit support forum. Originally posted on February 7, 2012; details may be outdated, see our docs for current behavior. Original thread
Hi guys,
I’ve been having a problem where it seems that you are throwing 500 errors when I check up on a resize request too soon after submitting it. Waiting 4 seconds doesn’t seem to be enough but 10 seconds seems to work. I’m not sure if its something I’m doing wrong or not. Can you see 500 errors on your side?
I try to pull an original image out of s3 and then crop out a part of it into another image. It does work if I wait long enough to poll about its status.
how do you check up on it? Do you check on the correct instance?
Are you using our jQuery plugin? It can fire callback functions when all sorts of things happen in our system. You should not need to manually poll the status of something.
Let me know in detail how you do it - possibly also point me to the url where you do it.
I use the python transloadit client to send a request to my transloadit template programmatically, my template in your system is: REDACTED
Then using the assembly id I get back I poll a url like this: http://api2.transloadit.com/assemblies/REDACTED until the assembly is finished. This is where the 500 error happens as far as I can tell. If I wait longer to poll it usually works.
yeah the problem is that this url only becomes available when we posted the status json already back to s3.
You need to access the url on the specific instance where the assembly is executed, for example http://api2.lisa.transloadit.com/assemblies/ in order to get the status out of the node process.
The assembly url that you get back should already include the instance name. Does it not do that? Maybe the (inofficial) python sdk has some magic to hide it?