CNDB-15508: Query planner metrics #2130
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Checklist before you submit for review
|
|
❌ Build ds-cassandra-pr-gate/PR-2130 rejected by Butler1 regressions found Found 1 new test failures
Found 1 known test failures |
ekaterinadimitrova2
left a comment
ekaterinadimitrova2
left a comment
There was a problem hiding this comment.
Not a full review, just started looking into this, just dropped a few quick comments
| public final Timer annGraphSearchLatency; | ||
|
|
||
| /** Query execution cost as estimated by the planner */ | ||
| public final Histogram costEstimated; |
There was a problem hiding this comment.
we probably need some parameter to be able to enable/disable these in case we overload a metrics server
This commit adds new metrics related to the operation of SAI query planner. The metrics should help checking if the query planner makes proper decisions by correlating them with the other metrics, e.g. the metrics of the actual query execution. Per-query metrics (histograms): - `RowsEstimated`: the estimated number of rows to be returned by the query - `CostEstimated`: the abstract cost of query execution - `LogSelectivityEstimated`: minus decimal logarithm of query selectivity, before applying the query LIMIT (0 means the query selects all rows, 5 means it selects 10^(-5) = 0.00001 subset of rows) - `IndexReferencesInQuery`: the number of index references in the unoptimized query execution plan (the same index may be referenced multiple times and counts separately) - `IndexReferencesInPlan`: the number of index references in the optimized query execution plan (the same index may be referenced multiple times and counts separately) Per-table: - `TotalRowsEstimated`: counts the sum of all row estimates from all completed queries - `TotalCostEstimated`: counts the sum of all cost estimates from all completed queries
pkolaczk
force-pushed
the
c15508-query-planner-metrics
branch
from
7986ec2 to
52e877c
Compare
adelapena
reviewed
Nov 18, 2025
Comment on lines
+328
to
+330
| if (originalPlan != null && optimizedPlan != null) { | ||
| queryPlanInfo = new QueryPlanInfo(originalPlan, optimizedPlan); | ||
| } else { |
| public final int indexReferencesInQuery; | ||
| public final int indexReferencesInPlan; | ||
|
|
||
| public QueryPlanInfo(@Nonnull Plan originalPlan, @Nonnull Plan optimizedPlan) |
There was a problem hiding this comment.
CNDB-15260 recently added SAI metrics to the log messages produced by the slow query logger. These messages contain most of the metrics that are tracked by QueryMonitorableExecutionInfo. The log messages also contain the query and the tree view of the query plan. I understand we don't need to add any of the metrics here to those log messages, because they don't add anything that is not already in the printed query and plan, is this correct?
This is how the log messages look like:
DEBUG [node1_isolatedExecutor:1] node1 2025-10-09 11:14:57,354 MonitoringTask.java:175 - 2 operations were slow in the last 131 msecs:
<SELECT * FROM distributed_test_keyspace.t WHERE (s = s_2 OR s = s_3) AND ALLOW FILTERING>, time 111 msec - slow timeout 100 msec
SAI slow query metrics:
sstablesHit: 2
segmentsHit: 2
partitionsRead: 2
rowsFiltered: 2
rowsPreFiltered: 0
trieSegmentsHit: 2
bkdPostingListsHit: 0
bkdSegmentsHit: 0
bkdPostingsSkips: 0
bkdPostingsDecodes: 0
triePostingsSkips: 0
triePostingsDecodes: 2
annGraphSearchLatencyNanos: 0
shadowedPrimaryKeyCount: 0
SAI slow query plan:
Limit 2147483647 (rows: 1.8, cost/row: 101.5, cost: 3000.0..3177.6)
└─ Filter (s = s_2 OR s = s_3) (sel: 1.000000000) (rows: 1.8, cost/row: 101.5, cost: 3000.0..3177.6)
└─ Fetch (rows: 1.8, cost/row: 101.5, cost: 3000.0..3177.6)
└─ Union (keys: 1.8, cost/key: 0.1, cost: 3000.0..3000.2)
├─ LiteralIndexScan of t_s_idx (sel: 0.250000000, step: 1.0) (keys: 1.0, cost/key: 0.1, cost: 1500.0..1500.1)
│ predicate: Expression{name: s, op: EQ, lower: (s_2, true), upper: (s_2, true), exclusions: []}
└─ LiteralIndexScan of t_s_idx (sel: 0.250000000, step: 1.0) (keys: 1.0, cost/key: 0.1, cost: 1500.0..1500.1)
predicate: Expression{name: s, op: EQ, lower: (s_3, true), upper: (s_3, true), exclusions: []}
| // Null means the query execution order hasn't been decided yet. | ||
| private FilterSortOrder filterSortOrder = null; | ||
| private Plan originalPlan = null; | ||
|
|
ekaterinadimitrova2
left a comment
ekaterinadimitrova2
left a comment
There was a problem hiding this comment.
Looks great, left just some questions.
There are some GH problems today and CI fails. You will need to restart it to check the results.
| // Determines the order of using indexes for filtering and sorting. | ||
| // Null means the query execution order hasn't been decided yet. | ||
| private FilterSortOrder filterSortOrder = null; | ||
| private Plan originalPlan = null; |
There was a problem hiding this comment.
Have you considered if storing only the metrics and not the full originalPlan and optimizedPlan may make more sense?
| import org.apache.cassandra.db.filter.IndexHints; | ||
| import org.apache.cassandra.db.filter.RowFilter; | ||
| import org.apache.cassandra.index.sai.IndexContext; | ||
| import org.apache.cassandra.index.sai.QueryContext; |
| { | ||
| Cost cost = cost(); | ||
| if (!(cost instanceof RowsIterationCost)) | ||
| throw new UnsupportedOperationException("Expected rows is only supported for plans returning rows (called on " + this.getClass() + ')'); |
There was a problem hiding this comment.
While this is true, shouldn't we return some 0 value maybe?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This commit adds new metrics related to the operation of SAI query
planner. The metrics should help checking if the query planner makes
proper decisions by correlating them with the other metrics,
e.g. the metrics of the actual query execution.
Per-query metrics (histograms):
RowsEstimated: the estimated number of rows to be returned bythe query
CostEstimated: the abstract cost of query executionInverseSelectivityEstimated: the inverse of query selectivity,before applying the query LIMIT
(1 means the query selects all rows, 10 means it
selects every 10th row, etc.)
IndexReferencesInQuery: the number of index references in theunoptimized query execution plan (the same index may
be referenced multiple times and counts separately)
IndexReferencesInPlan: the number of index references in theoptimized query execution plan (the same index may
be referenced multiple times and counts separately)
Per-table:
TotalRowsEstimated: counts the sum of all row estimates fromall completed queries
TotalCostEstimated: counts the sum of all cost estimates fromall completed queries
TotalQueriesCompletedInSelectivityGroup{N}, where N in [0, 12):counts the number of completed queries with selectivity S:
10^(-N-1) < S <= 10^(-N) for N < 11,
S <= 10^(-N) for N = 11
In other words, the higher the group N, the smaller fraction of rows
the query is estimated to return. The selectivity calculation
does not include the final LIMIT of the query.