feat: support state recovery when meta reboot

[yezizp2012]

What's changed and what's your intention?

As title, there are several changes in this PR to support state recovery when meta reboot:

• When meta leaves, frontend will re-subscribe until meta is online again. After re-subscribed, frontend will refresh its cache for catalog and worker info.
• Change the operation logic for force_stop_actors. When failover found in some compute nodes, other compute nodes that contains actors in related DAGs will panic and gone. Further more, the original implementation of force_stop_actors will cause panic in compute node, that's not acceptable when reuse it for meta reboot. Here we just inject a stop barrier for all exist actors in living compute nodes, that works and will help us to stop all exist actors.
• When meta reboot, we simply do a similar operation as recovery. This could be refine in the future when we have more barrier state persisted in meta store.

Checklist

• I have written necessary docs and comments
• I have added necessary unit tests and integration tests

Refer to a related PR or issue link (optional)

Resolve #1277

[codecov]

Codecov Report

Merging #1702 (1dce082) into main (1da4127) will increase coverage by 0.19%.
The diff coverage is 54.14%.

@@            Coverage Diff             @@
##             main    #1702      +/-   ##
==========================================
+ Coverage   71.15%   71.34%   +0.19%     
==========================================
  Files         598      599       +1     
  Lines       77556    77645      +89     
==========================================
+ Hits        55182    55399     +217     
+ Misses      22374    22246     -128     

Flag	Coverage Δ
rust	71.34% <54.14%> (+0.19%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/compute/src/rpc/service/stream_service.rs	0.00% <0.00%> (ø)
src/compute/src/server.rs	0.00% <0.00%> (ø)
src/ctl/src/common/meta_service.rs	0.00% <0.00%> (ø)
src/frontend/src/catalog/root_catalog.rs	71.96% <0.00%> (-2.25%)	⬇️
src/frontend/src/observer/observer_manager.rs	0.00% <0.00%> (ø)
src/frontend/src/scheduler/schedule.rs	9.58% <0.00%> (-0.56%)	⬇️
src/frontend/src/session.rs	44.55% <0.00%> (ø)
src/meta/src/model/mod.rs	99.05% <ø> (ø)
src/meta/src/rpc/server.rs	0.00% <0.00%> (ø)
src/rpc_client/src/meta_client.rs	0.00% <0.00%> (ø)
... and 22 more

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

[BugenZhao]

Rest LGTM. Great job!

[BugenZhao]

LGTM!