Fixing flaky endtoend/tablegc tests

Signed-off-by: Shlomi Noach <2607934+shlomi-noach@users.noreply.github.com>

.github/workflows/cluster_endtoend_tabletmanager_tablegc.yml

@@ -1,11 +1,11 @@
 # DO NOT MODIFY: THIS FILE IS GENERATED USING "make generate_ci_workflows"
 
-name: Cluster (27)
+name: Cluster (tabletmanager_tablegc)
 on: [push, pull_request]
 jobs:
 
   build:
-    name: Run endtoend tests on Cluster (27)
+    name: Run endtoend tests on Cluster (tabletmanager_tablegc)
     runs-on: ubuntu-18.04
 
     steps:
@@ -47,4 +47,4 @@ jobs:
       timeout-minutes: 30
       run: |
         source build.env
-        eatmydata -- go run test.go -docker=false -print-log -follow -shard 27
+        eatmydata -- go run test.go -docker=false -print-log -follow -shard tabletmanager_tablegc

test/ci_workflow_gen.go

@@ -51,7 +51,6 @@ var (
 		"23",
 		"24",
 		"26",
-		"27",
 		"vreplication_basic",
 		"vreplication_multicell",
 		"vreplication_cellalias",
@@ -65,6 +64,7 @@ var (
 		"onlineddl_singleton",
 		"tabletmanager_throttler",
 		"tabletmanager_throttler_custom_config",
+		"tabletmanager_tablegc",
 	}
 	// TODO: currently some percona tools including xtrabackup are installed on all clusters, we can possibly optimize
 	// this by only installing them in the required clusters

test/config.json

@@ -522,7 +522,7 @@
 			"Args": ["vitess.io/vitess/go/test/endtoend/tabletmanager/tablegc"],
 			"Command": [],
 			"Manual": false,
-			"Shard": "27",
+			"Shard": "tabletmanager_tablegc",
 			"RetryMax": 0,
 			"Tags": [
 				"site_test"