Archive appliance: pv disconnection/reconnection etc.

# Issue Description

As a developer wanting to use the archive appliance, I want to be sure that we handle PV disconnection/reconnection gracefully.

Currently:
- The AA has made an _architectural_ assumption that "not very many PVs are both being archived and disconnected". In fact, having too many disconnected PVs can completely lock up the archive appliance (as per their docs/source code). This is probably fine for accelerators but less good for us, so we may need to do some work around this
- We sometimes seem to be able to get the AA into a state where it loses connection to a PV but still thinks it's monitored, then it never retries the connection. This doesn't happen from pure EPICS tools (caget/camonitor etc) inside the container so I believe it's either an AA or a CAJ issue. In either case we need to look at it so that we can reliably reconnect to PVs when they reappear.

### How & Where?
- `isis-aa-config`
- `epicsarchiverap`

### Reproducible (Yes/No)?

Yes; start the archiver appliance and connect to 50+ pvs; most of them will connect successfully but some will not (different PVs each time, no discernible pattern, not consistent). The PVs which are disconnected do work from `caget`/`camonitor` inside the container.

### Additional Information

# Acceptance Criteria
- PV disconnection/reconnection in the AA works reliably
- Documented approach for how we'll get around AAs limitations around how many PVs can be disconnected

# How to Review
_{Before making a PR...}
- _Provide verbose instructions for the reviewer to test that your changes work and fix the issue_
- _Describe if/how you have implemented testing for this issue_
- _Provide screenshots of the feature to help the reviewer if relevant_

_{_If not applicable, write "Not applicable"_}

[comment]: <> (FILL IN BELOW)
> ...

_{**To the reviewer: Make sure to [update submodules](https://github.com/ISISComputingGroup/ibex_developers_manual/wiki/Git-workflow#reviewing-work-for-the-submodules-of-epics)!**}

Time in recording of planning meeting: 01h17m 2024/11/28