feat(dashboard): add backpressure to the relation dependency graph

[kwannoel]

I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.

What's changed and what's your intention?

Much of the code is borrowed from FragmentGraph and DependencyGraph components. Thanks for @xxchan and @BugenZhao who worked on that previously.

This PR adds a ddl level backpressure graph, to identify bottleneck MVs and Sinks in the stream graph.

From here, we can observe the specific fragment's backpressure graph by using the search box in the fragment graph page.

We also add a resetBackPressure button. This is for resetting embedded backpressures. Otherwise if there's a long idle period, followed by a sudden surge, the backpressure graph will be shown to be uncongested, since the average backpressure will include the idle period. By resetting the backpressure graph, we can restart the computation of average backpressure from the current point in time, and get a more accurate visualization.

Implementation details

The graph has relations, identified by relationId as its vertices.
And the edges are backpressures between the relations.

To render the vertices, we fetch all relation metadata from the meta-node.
The relation metadata includes the relation_id, schema_name, relation_name, as well as the parentIds.
relation_id and parentIds will be used to render the relation DAG, since they form the vertices of the graph.

Next, to render the edges, we have to remap our fragment level backpressure edges to relation level backpressure edges.

First, we obtain a mapping of input fragmentId and output fragmentId to their relation's ids.
An input fragmentId corresponds to a fragment which contains stream scan for instance.
An output fragmentId corresponds to a materialize fragment.

Next, our backpressure info contains a mapping of fragmentId-fragmentId to backpressure values.
We can map output fragment id to its corresponding relation, and likewise for input fragment id.
If it doesn't exist, we know that the edge can be removed, since it doesn't correspond to a relation-relation edge, and is internal to the relation itself.
From this we obtain a new mapping of relationId-relationId to backpressure values. Then we can use this to construct our backpressure edges for the ddl backpressure graph.

Checklist

• I have written necessary rustdoc comments
• I have added necessary unit tests and integration tests
• I have added test labels as necessary. See details.
• I have added fuzzing tests or opened an issue to track them. (Optional, recommended for new SQL features Sqlsmith: Sql feature generation #7934).
• My PR contains breaking changes. (If it deprecates some features, please create a tracking issue to remove them in the future).
• All checks passed in ./risedev check (or alias, ./risedev c)
• My PR changes performance-critical code. (Please run macro/micro-benchmarks and show the results.)

• My PR contains critical fixes that are necessary to be merged into the latest release. (Please check out the details)

Documentation

• My PR needs documentation updates. (Please use the Release note section below to summarize the impact on users)

Release note

If this PR includes changes that directly affect users or other significant modifications relevant to the community, kindly draft a release note to provide a concise summary of these changes. Please prioritize highlighting the impact these changes will have on users.

[kwannoel]

Merged both backpressure graph and dependency graph in the end.

Because there's certain design like:

• Labelling what is the ddl type, so we know what kind of output to expect (or to even expect any output)
• Accessing the fragments graph via the pop up.
• Accessing the mv catalog.

Which is helpful in getting more info about the backpressure stream job.

[xxchan]

Thank you. The UI looks good! (Didn't check backpressure computation logic.