Fix network object sharing race condition within the same process

[lyonsil]

This change is 

[lyonsil]

Prior to this change, we were not allowing a get and set (or two gets or two sets) for the same network object to run simultaneously in the same process. That creates a race condition where if one task calls get for network object X, then while waiting on the server to respond another task in the same process calls set for X, the set call will hang until the get times out.

After this change, we still prevent multiple gets or multiple sets for the same network object to run at the same time in the same process. This is desired so that we don't flood the network with duplicate calls or end up with orphaned data structures that get overwritten with duplicates in maps. However, we will now allow a get and a set to run simultaneously for the same object.

Reviewable status: 0 of 1 files reviewed, all discussions resolved

[tjcouch-sil]

thank you!!

Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved