missing tombstones in dse 6.0

CHANGELOG.txt

@@ -1,3 +1,7 @@
+sperf 0.6.16
+------------
+* tombstone warnings for 6.0.x are not parsed correctly
+
 sperf 0.6.15
 ------------
 * added detection for indexing backoff with auto soft commit

pysper/__init__.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """top level module for sperf python port"""
-VERSION = "0.6.15"
+VERSION = "0.6.16"

pysper/changelog.py

@@ -1,4 +1,8 @@
 CHANGES = """
+sperf 0.6.16
+------------
+* tombstone warnings for 6.0.x are not parsed correctly
+
 sperf 0.6.15
 ------------
 * added detection for indexing backoff with auto soft commit

pysper/commands/core/diag.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-"""cassread command wiring"""
+"""sperf core diag command wiring"""
 from pysper.commands import flags
 from pysper import VERSION
 from pysper.core.diag import parse_diag, generate_report
@@ -72,7 +72,7 @@ def add_args(diag_parser):
 
 
 def add_flags(subparsers, name, run_func, is_deprecated=False):
-    """so we can share the parser with the legacy sperf cassread command"""
+    """so we can share the parser with the legacy 'sperf core diag' command"""
     help_text = "Generates a diagtarball report. DSE 5.0-6.8"
     if is_deprecated:
         help_text = help_text + ". DEPRECATED: use 'sperf core diag' instead"
@@ -84,7 +84,7 @@ def add_flags(subparsers, name, run_func, is_deprecated=False):
 
 
 def build(subparsers):
-    """builds the cassread command up, assumes argparser api"""
+    """builds the sperf core diag command up, assumes argparser api"""
     add_flags(subparsers, "diag", run)
 
 

pysper/commands/sperf.py

@@ -98,7 +98,7 @@ def run():
             print(str(ex))
             print("")
             print(
-                "To show strack trace use the -v flag. For example: 'sperf -v cassread'"
+                "To show strack trace use the -v flag. For example: 'sperf -v core statuslogger'"
             )
     else:
         print("sperf version %s" % VERSION)

pysper/commands/sperf_default.py

@@ -33,4 +33,6 @@ def run(args):
             raise ex
         print(str(ex))
         print("")
-        print("To show strack trace use the -v flag. For example: 'sperf -v cassread'")
+        print(
+            "To show strack trace use the -v flag. For example: 'sperf -v core statuslogger'"
+        )

pysper/core/diag/__init__.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-"""information collection and report writing for the cassread tool"""
+"""information collection and report writing for the sperf core diag command"""
 from pysper.core.diag import reporter, node_env, table_stats, config_diff, read_ahead
 from pysper import diag, util
 

pysper/core/diag/reporter.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-"""formats the output of the cassread parsed data"""
+"""formats the output of the sperf core diag parsed data"""
 from pysper import humanize, env
 
 # used for report formatting

pysper/parser/cases.py

@@ -699,6 +699,18 @@ def zc_rules():
 def tombstone_rules():
     """catch tombstone problems"""
     return (
+        case("ReadCommand"),
+        rule(
+            capture(
+                r"Read (?P<live_rows>[0-9]*) live rows and (?P<tombstones>[0-9]*) tombstone cells for query (?P<query>.*) \(see tombstone_warn_threshold\)"
+            ),
+            convert(int, "tombstones"),
+            update(
+                event_product="tombstone",
+                event_category="reading",
+                event_type="tpc_scan_warn",
+            ),
+        ),
         case("MessageDeliveryTask"),
         rule(
             capture(

scripts/sperf

@@ -17,9 +17,10 @@
 import sys
 import os
 import platform
-if platform.python_version() < '3.7':
+
+if platform.python_version() < "3.7":
     raise Exception("sperf requires at least Python 3.7, consider a packaged release")
-sys.path.append(os.path.join(os.path.dirname(os.path.abspath(__file__)), '..'))
+sys.path.append(os.path.join(os.path.dirname(os.path.abspath(__file__)), ".."))
 from pysper.commands import sperf
 
 if __name__ == "__main__":

setup.py

@@ -8,7 +8,7 @@
 
 setup(
     name="sperf",
-    version="0.6.15",
+    version="0.6.16",
     description="Diagnostic utility for DSE and Cassandra",
     url="https://www.github.com/DataStax-Toolkit/sperf",
     scripts=["scripts/sperf"],

tests/commands/test_sperf.py

@@ -47,7 +47,7 @@ def test_sperf_will_break_if_incorrect_comamnd_is_used(self):
         self.assertTrue(success)
 
     def test_core_diag_wired(self):
-        """verify cassread is connected"""
+        """verify sperf core diag is connected"""
         parser = sperf.build_parser()
         args = parser.parse_args(["core", "diag"])
         self.assertTrue(hasattr(args, "func"))

tests/core/test_statuslogger.py

@@ -45,7 +45,7 @@ def test_skip_duplicate_events_diag(self):
         self.assertTrue(sl.analyzed)
         self.assertEqual(len(sl.nodes), 3)
         s = Summary(sl.nodes)
-        self.assertEqual(s.lines, 22054)
+        self.assertEqual(s.lines, 22055)
         self.assertEqual(s.skipped_lines, 445)
         self.assertEqual(
             s.get_busiest_stages()[0],

tests/test_sperf_default.py

@@ -59,7 +59,8 @@ def run():
 ---------------
 * There were 16 incidents of GC over 500ms. Run `sperf core gc` for more analysis.
 * There were drops of the following request types: MUTATION for a total of 1096 drops. Run sperf core statuslogger and look for high pending stages for those messages types.
-* Global local backpressure was active on the following nodes: 10.101.35.102 (1 times). GC was over target %s times however despire TPC backpressure being active it may be dangerous to raise TPC limits, run sperf core statuslogger for further analysis on the type of requests that are pending.""",
+* Global local backpressure was active on the following nodes: 10.101.35.102 (1 times). GC was over target %s times however despire TPC backpressure being active it may be dangerous to raise TPC limits, run sperf core statuslogger for further analysis on the type of requests that are pending.
+* Tombstone warnings found, there were 1951 total tombstones scanned. The data model has a problem.""",
         )
 
     def test_sperf_68(self):

tests/test_statuslogger.py

@@ -47,7 +47,7 @@ def run():
             output,
             "sperf core statuslogger version: %s\n" % (VERSION)
             + """
-Summary (22,054 lines)
+Summary (22,055 lines)
 Summary (445 skipped lines)
 
 dse versions: {'6.7.7'}

tests/testdata/diag/DSE_CLUSTER/nodes/10.101.33.205/logs/cassandra/system.log

@@ -2462,3 +2462,4 @@ INFO  [GCInspector:1] 2020-01-10 17:19:08,323  GCInspector.java:313 - G1 Young G
 INFO  [commitScheduler-6-thread-1] 2020-01-10 17:19:14,043  DocumentsWriter.java:785 - Waited 16 milliseconds for 7 flushing segments.
 INFO  [commitScheduler-6-thread-1] 2020-01-10 17:19:24,124  DocumentsWriter.java:785 - Waited 0 milliseconds for 0 flushing segments.
 INFO  [NodeSyncMaintenanceTasks:1] 2020-01-10 17:19:43,531  NodeSyncMaintenanceTasks.java:172 - In last 10m: validated 0B (0B/s), 0% was inconsistent.
+WARN  [CoreThread-3] 2020-01-10 17:19:47,591  ReadCommand.java:535 - Read 1 live rows and 1951 tombstone cells for query SELECT * FROM keyspace1.standard1 WHERE id = 894e165d-beba-433d-ab52-055c570fcb9b LIMIT 2000 (see tombstone_warn_threshold)