[+] enhance Query Performance Analysis pg dashboard

[0xgouda]

• Fix the panel queries' logic to use the sum(col - col_lag) pattern used in other dashboards instead of max(col) - min(col), which skips counting the changes between the elements at the boundaries of each time segment.
• Fix the panel queries' logic to show the top $n for each time segment, not across the entire user-specified range.
• Address Problem in query logic of some panels #1152 for this dashboard
• Change the panels to have a stacked view, which is helpful for deducing the cumulative load of the top $n queries
• Make panels display the rate per call/second instead of raw value increase (similar to prometheus rate()).

Screenshots (the updated version is hosted on the dc1 vm and is probably better to inspect it from there) :

---

TODO:

• ask for review by our dbas.
• evaluate how slow the new queries might be on full 2-weeks data, for both timescaledb and vanilla postgresql setups.

[coveralls]

Coverage Report for CI Build 24842665207

Coverage decreased (-0.2%) to 83.333%

Details

• Coverage decreased (-0.2%) from the base build.
• Patch coverage: No coverable lines changed in this PR.
• 8 coverage regressions across 1 file.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

8 previously-covered lines in 1 file lost coverage.

File	Lines Losing Coverage	Coverage
internal/sinks/prometheus.go	8	93.95%

---

Coverage Stats

Relevant Lines:	5322
Covered Lines:	4435
Line Coverage:	83.33%
Coverage Strength:	0.95 hits per line

---

💛 - Coveralls

[pashagolub]

Looks great!