[teraslice-messaging] fix memory leak caused by abortController

[busma13]

This PR makes the following changes:
-bump teraslice-messaging from 1.10.3 to 1.10.4

• Remove the abortController from the execution-controller client
• Create an abortController in the core client handleSendResponse function if sendAbortSignal is set to true. This is only the case for when the execution-controller client sends the worker:slice:complete event. On a server:shutdown event the abortController will abort awaiting a server response so it can shutdown properly.

ref: #3945

[busma13]

Using the promMetrics exporter I was able to see the resident memory of the worker increasing from ~225MB at. first scrape to ~367MB after 48000 slices on the current master branch. Using this branch the memory increased from ~230MB at first scrape to ~232MB after 72000 slices.

[godber]

Unfortunately I didn't catch until now that the teraslice version itself was not bumped. Doesn't bumping teraslice-messaging result in teraslice being bumped?