support dumping metrics by instance

[zyguan]

What problem does this PR solve?

For a large cluster, getSeriesNum may always fail due to quota limit. As a result, significant metrics like tidb_tikvclient_request_seconds and tikv_grpc_msg_duration_seconds are often missing, which blocks troubleshooting.

What is changed and how it works?

With this PR, if getSeriesNum failed, diag will fallback to by-instance dump mode, which get instance list first and call collectMetric by instance.

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No code

Code changes

• Has exported function/method change
• Has exported variable/fields change
• Has interface methods change
• Has persistent data change

Side effects

• Possible performance regression
• Increased code complexity
• Breaking backward compatibility

Related changes

• Need to cherry-pick to the release branch
• Need to update the documentation

[CLAassistant]

All committers have signed the CLA.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (4c0b393) 30.21% compared to head (913f819) 30.21%.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #470   +/-   ##
=======================================
  Coverage   30.21%   30.21%           
=======================================
  Files          52       52           
  Lines        4416     4416           
=======================================
  Hits         1334     1334           
  Misses       2911     2911           
  Partials      171      171           

Flag	Coverage Δ
unittests	30.21% <ø> (ø)

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.