[ez][HUD] Remove rockset from various comments, remove enableClickhouse function

torchci/components/benchmark/BranchAndCommitPicker.tsx

@@ -14,7 +14,7 @@ import useSWR from "swr";
 
 // Keep the mapping from workflow ID to commit, so that we can use it to
 // zoom in and out of the graph. NB: this is to avoid sending commit sha
-// again from Rockset in the compilers_benchmark_performance query which
+// again from the database in the compilers_benchmark_performance query which
 // already returns close to the 6MB data transfer limit. I need to figure
 // out a way to compress the data later
 export const COMMIT_TO_WORKFLOW_ID: { [k: string]: number } = {};

torchci/lib/clickhouse.ts

@@ -78,8 +78,3 @@ export async function queryClickhouseSaved(
     Object.fromEntries(queryParams)
   );
 }
-
-export function enableClickhouse() {
-  // Use this to quickly toggle between clickhouse and rockset
-  return process.env.USE_CLICKHOUSE == "true";
-}

torchci/lib/fetchFlakyTests.ts

@@ -106,11 +106,12 @@ where
     numHours,
   });
 
-  // For every failed test, query rockset for jobs that had file level reruns of
+  // For every failed test, query the database for jobs that had file level reruns of
   // the test in the past numHours.  Do this separately because a join on
   // test_run_s3 takes a long time.  Batch the query since rockset doesn't allow
   // more tha 150 concurrent queries.  Flatten the accumulator since it ends up
   // being an array of arrays.
+  // TODO: Check if batching is still needed now that we are using clickhouse
   let rerunTestsUnflattened: any[] = [];
   for (let i = 0; i < failedTestsResults.length; i += 25) {
     rerunTestsUnflattened.push(

torchci/lib/metricUtils.ts

@@ -53,7 +53,7 @@ function increaseFailureCount(
 }
 
 export function approximateSuccessByJobName(
-  // The data from Rockset is sorted by time DESC, so newer commits come first
+  // The data is sorted by time DESC, so newer commits come first
   data?: JobsPerCommitData[]
 ) {
   const successesByJobName: { [success: string]: number } = {};
@@ -84,7 +84,7 @@ export function approximateSuccessByJobName(
 }
 
 export function approximateFailureByType(
-  // The data from Rockset is sorted by time DESC, so newer commits come first
+  // The data is sorted by time DESC, so newer commits come first
   data?: JobsPerCommitData[],
   broken_trunk_threshold: number = BROKEN_TRUNK_THRESHOLD,
   outage_threshold: number = OUTAGE_THRESHOLD
@@ -154,7 +154,7 @@ export function approximateFailureByType(
 }
 
 export function approximateFailureByTypePercent(
-  // The data from Rockset is sorted by time DESC, so newer commits come first
+  // The data is sorted by time DESC, so newer commits come first
   data?: JobsPerCommitData[],
   broken_trunk_threshold: number = BROKEN_TRUNK_THRESHOLD,
   outage_threshold: number = OUTAGE_THRESHOLD

torchci/lib/types.ts

@@ -1,5 +1,5 @@
 /**
- * Represents the individual job information returned by Rockset.
+ * Represents the individual job information returned by queries.
  */
 export interface BasicJobData {
   name?: string;
@@ -37,7 +37,7 @@ export interface RecentWorkflowsData extends BasicJobData {
   // only included if this is a job and not a workflow, if it is a workflow, the name is in the name field
   name: string; // In BasicJobData, but required here
   workflowId: number;
-  // Each workflow file has an id. In rockset this is workflow_run.workflow_id.
+  // Each workflow file has an id. In the webhook this is workflow_id.
   // This can be used to group normal workflows (ex trunk) and those that failed
   // to run (ex .github/workflows/trunk.yml) together even when they have
   // different names.