Add --use-known-results-for-testing flag to supercluster. Use helm te…

…mplating to conditionally include CATCHUP_SKIP_KNOWN_RESULTS_FOR_TESTING in PubnetParallelCatchupV2 workers stellar-core.cfg when this flag is provided.
stellar · Dec 6, 2024 · 2c8c0bc · 2c8c0bc
1 parent edf62fb
commit 2c8c0bc
Show file tree

Hide file tree

Showing 6 changed files with 22 additions and 7 deletions.
diff --git a/src/App/Program.fs b/src/App/Program.fs
@@ -109,7 +109,8 @@ type MissionOptions
         pubnetParallelCatchupEndLedger: int option,
         pubnetParallelCatchupNumWorkers: int,
         tag: string option,
-        numRuns: int option
+        numRuns: int option,
+        catchupSkipKnownResultsForTesting: bool
     ) =
 
     [<Option('k', "kubeconfig", HelpText = "Kubernetes config file", Required = false, Default = "~/.kube/config")>]
@@ -458,6 +459,12 @@ type MissionOptions
              Required = false)>]
     member self.NumRuns = numRuns
 
+    [<Option("catchup-skip-known-results-for-testing",
+             HelpText = "when this flag is provided, pubnet parallel catchup workers will run with CATCHUP_SKIP_KNOWN_RESULTS_FOR_TESTING = true, resulting in skipping application of failed transaction and signature verification",
+             Required = false,
+             Default = false)>]
+    member self.CatchupSkipKnownResultsForTesting = catchupSkipKnownResultsForTesting
+
 let splitLabel (lab: string) : (string * string option) =
     match lab.Split ':' with
     | [| x |] -> (x, None)
@@ -571,7 +578,8 @@ let main argv =
                   pubnetParallelCatchupNumWorkers = 128
                   tag = None
                   numRuns = None
-                  enableTailLogging = true }
+                  enableTailLogging = true
+                  catchupSkipKnownResultsForTesting = false }
 
             let nCfg = MakeNetworkCfg ctx [] None
             use formation = kube.MakeEmptyFormation nCfg
@@ -707,7 +715,8 @@ let main argv =
                                pubnetParallelCatchupNumWorkers = mission.PubnetParallelCatchupNumWorkers
                                tag = mission.Tag
                                numRuns = mission.NumRuns
-                               enableTailLogging = true }
+                               enableTailLogging = true
+                               catchupSkipKnownResultsForTesting = mission.CatchupSkipKnownResultsForTesting }
 
                          allMissions.[m] missionContext
 

diff --git a/src/FSLibrary.Tests/Tests.fs b/src/FSLibrary.Tests/Tests.fs
@@ -117,7 +117,8 @@ let ctx : MissionContext =
       pubnetParallelCatchupNumWorkers = 128
       tag = None
       numRuns = None
-      enableTailLogging = true }
+      enableTailLogging = true
+      catchupSkipKnownResultsForTesting = false }
 
 let netdata = __SOURCE_DIRECTORY__ + "/../../../data/public-network-data-2024-08-01.json"
 let pubkeys = __SOURCE_DIRECTORY__ + "/../../../data/tier1keys.json"

diff --git a/src/FSLibrary/MissionHistoryPubnetParallelCatchupV2.fs b/src/FSLibrary/MissionHistoryPubnetParallelCatchupV2.fs
@@ -67,6 +67,7 @@ let installProject (context: MissionContext) =
     setOptions.Add(sprintf "worker.stellar_core_image=%s" context.image)
     setOptions.Add(sprintf "worker.replicas=%d" context.pubnetParallelCatchupNumWorkers)
     setOptions.Add(sprintf "range_generator.params.starting_ledger=%d" context.pubnetParallelCatchupStartingLedger)
+    setOptions.Add(sprintf "worker.catchup_skip_known_results_for_testing=%b" context.catchupSkipKnownResultsForTesting)
 
     let endLedger =
         match context.pubnetParallelCatchupEndLedger with

diff --git a/src/FSLibrary/StellarMissionContext.fs b/src/FSLibrary/StellarMissionContext.fs
@@ -109,4 +109,5 @@ type MissionContext =
       // and SimulatePubnet to fail on the heartbeat handler due to what looks like a
       // server disconnection. Our solution for now is to just disable tail logging on
       // those missions.
-      enableTailLogging: bool }
+      enableTailLogging: bool
+      catchupSkipKnownResultsForTesting: bool }
diff --git a/src/MissionParallelCatchup/parallel_catchup_helm/files/worker.sh b/src/MissionParallelCatchup/parallel_catchup_helm/files/worker.sh
@@ -29,8 +29,7 @@ if [ -n "$JOB_KEY" ]; then
     # Start timer
     START_TIME=$(date +%s)  
     echo "Processing job: $JOB_KEY"
-
-    if [ ! "$(/usr/bin/stellar-core --conf /config/stellar-core.cfg new-db --console&&
+   if [ ! "$(/usr/bin/stellar-core --conf /config/stellar-core.cfg new-db --console&&
     /usr/bin/stellar-core --conf /config/stellar-core.cfg catchup "$JOB_KEY" --metric 'ledger.transaction.apply' --console)" ]; then
     echo "Error processing job: $JOB_KEY"
     redis-cli -h "$REDIS_HOST" -p "$REDIS_PORT" LPUSH "$FAILED_QUEUE" "$JOB_KEY|$POD_NAME" # enhance the entry with pod name for tracking

diff --git a/src/MissionParallelCatchup/parallel_catchup_helm/templates/catchup_workers.yaml b/src/MissionParallelCatchup/parallel_catchup_helm/templates/catchup_workers.yaml
@@ -102,6 +102,10 @@ metadata:
   name: stellar-core-config
 data:
   stellar-core.cfg: |-
+    {{- if .Values.worker.catchup_skip_known_results_for_testing }}
+    {{ "\n" }}
+    CATCHUP_SKIP_KNOWN_RESULTS_FOR_TESTING = true
+    {{- end }}
     {{- (.Files.Get "files/stellar-core.cfg") | nindent 4 }}
 ---
 apiVersion: v1