cassandra-pr mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [cassandra] dcapwell commented on a change in pull request #575: CASSANDRA-15776 port nodetool_test.py test_refresh_size_estimates_clears_invalid_entries to jvm dtest and fixed bug where size_estimates isn't used for clearing tables
Date Thu, 07 May 2020 17:32:15 GMT

dcapwell commented on a change in pull request #575:
URL: https://github.com/apache/cassandra/pull/575#discussion_r421674154



##########
File path: src/java/org/apache/cassandra/db/SystemKeyspace.java
##########
@@ -1344,12 +1344,17 @@ public static void clearEstimates(String keyspace)
     public static synchronized SetMultimap<String, String> getTablesWithSizeEstimates()
     {
         SetMultimap<String, String> keyspaceTableMap = HashMultimap.create();
-        String cql = String.format("SELECT keyspace_name, table_name FROM %s", TableEstimates.toString(),
TABLE_ESTIMATES_TYPE_PRIMARY);
-        UntypedResultSet rs = executeInternal(cql);
-        for (UntypedResultSet.Row row : rs)
+        // Its possible that size_estimates knows about a different set of keyspace/tables
than table_estimates (mostly
+        // caused by external systems modifying the tables, such as dtest) so query both
+        for (String cql : Arrays.asList(
+            "SELECT keyspace_name, table_name FROM " + TableEstimates.toString(),
+            "SELECT keyspace_name, table_name FROM " + LegacySizeEstimates.toString()))

Review comment:
       > Do we expect the set of keyspaces in table_estimates and size_estimates to be
mutually exclusive? If not, aren't you clobbering the estimates from table_estimates with
the values from size_estimates table? Is that the expected behavior?
   
   @dineshjoshi I expect them to conflict 99% of the time (1% for random users writing random
data), which is why a set is used to dedup; after a reboot they should have identical tables.
   
   >  clobbering the estimates from table_estimates with the values from size_estimates
table?
   
   This function only looks for table names, not values, so no clobbering estimates.




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org


Mime
View raw message