Add configurable cluster startup timeout and verbose output (#242)

* Add configurable cluster startup timeout and verbose output

- Add --cluster-start-timeout CLI argument (default 40 seconds)
- Add verbose progress output during cluster startup:
  - Show shard startup progress (Started X/Y shards)
  - Show cluster topology configuration progress
  - Show cluster wait progress with shard OK/slots agreement counts
- Use CLUSTER ADDSLOTSRANGE instead of ADDSLOTS for efficiency
- Add error logging for slot assignment and shard startup failures

* address PR review

* validate cluster-start-timeout at argument parsing time

* remove redundant validation from waitCluster

* fix: only pass clusterStartTimeout to ClusterEnv

Author: fcostaoliveira

RLTest/__main__.py

@@ -148,6 +148,11 @@ def do_normal_conn(self, line):
     '--cluster_node_timeout', default=5000,
     help='sets the node timeout on cluster in milliseconds')
 
+parser.add_argument(
+    '--cluster-start-timeout', default=40, type=int,
+    help='timeout in seconds to wait for cluster to be ready (default 40 seconds). '
+         'Increase for large shard counts (e.g., 99 shards).')
+
 parser.add_argument(
     '--cluster_credentials',
     help='enterprise cluster cluster_credentials "username:password", relevent only when running with cluster_existing-env')
@@ -538,6 +543,9 @@ def __init__(self):
         Defaults.tls_passphrase = self.args.tls_passphrase
         Defaults.oss_password = self.args.oss_password
         Defaults.cluster_node_timeout = self.args.cluster_node_timeout
+        Defaults.cluster_start_timeout = self.args.cluster_start_timeout
+        if Defaults.cluster_start_timeout < 5:
+            raise Exception('--cluster-start-timeout must be at least 5 seconds')
         Defaults.enable_debug_command = True if self.args.allow_unsafe else self.args.enable_debug_command
         Defaults.enable_protected_configs = True if self.args.allow_unsafe else self.args.enable_protected_configs
         Defaults.enable_module_command = True if self.args.allow_unsafe else self.args.enable_module_command

RLTest/env.py

@@ -142,6 +142,7 @@ class Defaults:
     randomize_ports = False
     oss_password = None
     cluster_node_timeout = None
+    cluster_start_timeout = 40
     curr_test_name = None
     port = 6379
     enable_debug_command = False
@@ -315,6 +316,7 @@ def getEnvByName(self):
                                         **kwargs)
         if self.env == 'oss-cluster':
             kwargs['password'] = Defaults.oss_password if self.password is None else self.password
+            kwargs['clusterStartTimeout'] = Defaults.cluster_start_timeout
             return ClusterEnv(shardsCount=self.shardsCount, redisBinaryPath=self.redisBinaryPath,
                               outputFilesFormat='%s-' + '%s-oss-cluster' % test_fname,
                               randomizePorts=Defaults.randomize_ports,
@@ -369,7 +371,7 @@ def getEnvKwargs(self):
             'terminateRetries': self.terminateRetries,
             'terminateRetrySecs': self.terminateRetrySecs,
             'redisConfigFile': self.redisConfigFile,
-            'dualTLS': self.dualTLS
+            'dualTLS': self.dualTLS,
         }
         return kwargs
 

RLTest/redis_cluster.py

@@ -6,6 +6,9 @@
 import time
 from RLTest.utils import Colors
 
+# Interval in seconds between status updates during cluster wait
+CLUSTER_STATUS_INTERVAL_SEC = 5
+
 
 class ClusterEnv(object):
     def __init__(self, **kwargs):
@@ -23,6 +26,7 @@ def __init__(self, **kwargs):
         self.protocol = kwargs.get('protocol', 2)
         self.terminateRetries = kwargs.get('terminateRetries', None)
         self.terminateRetrySecs = kwargs.get('terminateRetrySecs', None)
+        self.clusterStartTimeout = kwargs.pop('clusterStartTimeout', 40)
         startPort = kwargs.pop('port', 10000)
         totalRedises = self.shardsCount * (2 if useSlaves else 1)
         randomizePorts = kwargs.pop('randomizePorts', False)
@@ -50,7 +54,8 @@ def getInformationBeforeDispose(self):
     def getInformationAfterDispose(self):
         return [shard.getInformationAfterDispose() for shard in self.shards]
 
-    def _agreeOk(self):
+    def _countOk(self):
+        """Returns count of shards reporting cluster_state:ok"""
         ok = 0
         for shard in self.shards:
             con = shard.getConnection()
@@ -61,9 +66,10 @@ def _agreeOk(self):
                 continue
             if 'cluster_state:ok' in str(status):
                 ok += 1
-        return ok == len(self.shards)
+        return ok
 
-    def _agreeSlots(self):
+    def _countAgreeSlots(self):
+        """Returns count of shards that agree on slots view"""
         ok = 0
         first_view = None
         for shard in self.shards:
@@ -77,35 +83,63 @@ def _agreeSlots(self):
                 first_view = slots_view
             if slots_view == first_view:
                 ok += 1
-        return ok == len(self.shards)
+        return ok
 
-    def waitCluster(self, timeout_sec=40):
+    def waitCluster(self, timeout_sec=40, verbose=True):
         st = time.time()
+        last_status_time = st
+        total_shards = len(self.shards)
+
+        if verbose:
+            print(Colors.Yellow('Waiting for cluster to be ready (timeout: %d seconds, %d shards)...' %
+                                (timeout_sec, total_shards)))
 
         while st + timeout_sec > time.time():
-            if self._agreeOk() and self._agreeSlots():
+            ok_count = self._countOk()
+            slots_count = self._countAgreeSlots()
+
+            if ok_count == total_shards and slots_count == total_shards:
+                elapsed = time.time() - st
+                if verbose:
+                    print(Colors.Green('Cluster is ready after %.1f seconds' % elapsed))
                 for shard in self.shards:
                     try:
                         shard.getConnection().execute_command('SEARCH.CLUSTERREFRESH')
                     except Exception:
                         pass
                 return
 
+            # Print periodic status update
+            now = time.time()
+            if verbose and (now - last_status_time) >= CLUSTER_STATUS_INTERVAL_SEC:
+                elapsed = now - st
+                print(Colors.Yellow('  Cluster wait: %.1fs elapsed - %d/%d shards OK, %d/%d agree on slots...' %
+                                    (elapsed, ok_count, total_shards, slots_count, total_shards)))
+                last_status_time = now
+
             time.sleep(0.1)
         raise RuntimeError(
             "Cluster OK wait loop timed out after %s seconds" % timeout_sec)
 
     def startEnv(self, masters=True, slaves=True):
         if self.envIsUp == True:
             return  # env is already up
+
+        total_shards = len(self.shards)
+        print(Colors.Yellow('Starting cluster with %d shards...' % total_shards))
+
         try:
-            for shard in self.shards:
+            for i, shard in enumerate(self.shards):
                 shard.startEnv(masters, slaves)
-        except Exception:
+                print(Colors.Yellow('  Started shard %d/%d' % (i + 1, total_shards)))
+        except Exception as e:
+            print(Colors.Bred('Error starting shard %d: %s' % (i + 1, str(e))))
+            print(Colors.Bred('Stopping all shards...'))
             for shard in self.shards:
                 shard.stopEnv()
             raise
 
+        print(Colors.Yellow('Configuring cluster topology...'))
         slots_per_node = int(16384 / len(self.shards)) + 1
         for i, shard in enumerate(self.shards):
             con = shard.getConnection()
@@ -121,10 +155,14 @@ def startEnv(self, masters=True, slaves=True):
             try:
                 con.execute_command('CLUSTER', 'ADDSLOTS', *(str(x)
                                     for x in range(start_slot, end_slot)))
-            except Exception:
-                pass
+            except Exception as e:
+                print(Colors.Bred('  Error assigning slots %d-%d to shard %d: %s' %
+                                  (start_slot, end_slot - 1, i + 1, str(e))))
 
-        self.waitCluster()
+            print(Colors.Yellow('  Configured shard %d/%d (slots %d-%d)' %
+                                (i + 1, total_shards, start_slot, min(end_slot - 1, 16383))))
+
+        self.waitCluster(timeout_sec=self.clusterStartTimeout)
         self.envIsUp = True
         self.envIsHealthy = True