New cache implementation for the server

[ljeub-pometry]

What changes were proposed in this pull request?

• new cache based on quick_cache
• no more evicting of dirty graphs
• atomic replace of cached graphs
• specifying port=0 for server start now works
• starting a server without specifying a port now falls back to using an arbitrary port if the default port is in use
• python tests now work even if there is a graphql server running on 1736
• server startup no longer pings the network interface to figure out if the server has started (it uses messaging between threads instead).

Why are the changes needed?

• race conditions in the server when replacing graphs

Does this PR introduce any user-facing change? If yes is this documented?

How was this patch tested?

• improved the tests such that they fail when the flush on drop fails (but only when compiled with a special test feature)

Are there any further changes required?

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Rust Benchmark'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 2.

Benchmark suite	Current: a97712a	Previous: 9823ef7	Ratio
lotr_graph/num_edges	3 ns/iter (± 0)	0 ns/iter (± 0)	+∞
lotr_graph/num_nodes	3 ns/iter (± 0)	1 ns/iter (± 0)	3
lotr_graph/graph_latest	2 ns/iter (± 0)	0 ns/iter (± 0)	+∞
lotr_graph_materialise/materialize	6113640 ns/iter (± 8981)	1564816 ns/iter (± 35303)	3.91
lotr_graph_window_100/num_nodes	11 ns/iter (± 0)	5 ns/iter (± 0)	2.20
lotr_graph_window_100/iterate_exploded_edges	670769 ns/iter (± 1214)	325242 ns/iter (± 847)	2.06
lotr_graph_window_100_materialise/materialize	6526413 ns/iter (± 19681)	1669150 ns/iter (± 10700)	3.91
lotr_graph_window_10/iterate nodes	26767 ns/iter (± 40)	11339 ns/iter (± 40)	2.36
lotr_graph_window_10/iterate_exploded_edges	326315 ns/iter (± 540)	155788 ns/iter (± 1001)	2.09
lotr_graph_window_10_materialise/materialize	2815635 ns/iter (± 6382)	971980 ns/iter (± 4278)	2.90
lotr_graph_subgraph_10pc_materialise/materialize	1279120 ns/iter (± 12799)	334634 ns/iter (± 1287)	3.82
lotr_graph_subgraph_10pc_windowed/has_node_existing	125 ns/iter (± 8)	62 ns/iter (± 14)	2.02
lotr_graph_subgraph_10pc_windowed/iterate nodes	4518 ns/iter (± 71)	1365 ns/iter (± 3)	3.31
lotr_graph_subgraph_10pc_windowed_materialise/materialize	785861 ns/iter (± 8734)	230399 ns/iter (± 2617)	3.41
lotr_graph_window_50_layered/has_node_existing	347 ns/iter (± 18)	129 ns/iter (± 12)	2.69
lotr_graph_window_50_layered/iterate nodes	62557 ns/iter (± 515)	19308 ns/iter (± 47)	3.24
lotr_graph_window_50_layered_materialise/materialize	26422755 ns/iter (± 165109)	3488825 ns/iter (± 24948)	7.57
lotr_graph_persistent_window_50_layered/num_edges_temporal	541196 ns/iter (± 5663)	192686 ns/iter (± 1569)	2.81
lotr_graph_persistent_window_50_layered/has_node_existing	393 ns/iter (± 289)	174 ns/iter (± 83)	2.26
lotr_graph_persistent_window_50_layered/iterate nodes	86422 ns/iter (± 1453)	35886 ns/iter (± 191)	2.41
lotr_graph_persistent_window_50_layered/iterate_exploded_edges	3880521 ns/iter (± 7594)	1659940 ns/iter (± 19402)	2.34
lotr_graph_persistent_window_50_layered_materialise/materialize	48034811 ns/iter (± 121405)	5298035 ns/iter (± 147912)	9.07
lotr_graph/proto_encode	7398617 ns/iter (± 46165)	1157897 ns/iter (± 73709)	6.39

This comment was automatically generated by workflow using github-action-benchmark.