phoenix-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Samarth Jain (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (PHOENIX-4151) Tests extending BaseQueryIT are flapping
Date Sat, 02 Sep 2017 17:18:00 GMT

    [ https://issues.apache.org/jira/browse/PHOENIX-4151?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16151544#comment-16151544
] 

Samarth Jain commented on PHOENIX-4151:
---------------------------------------

There is some data in the data table on which this index is being created on. 

{code}
this.tableName =
                    initATableValues(generateUniqueName(), tenantId, getDefaultSplits(tenantId),
                        date = new Date(System.currentTimeMillis()), ts, getUrl(), tableDDLOptions);

props.setProperty(PhoenixRuntime.CURRENT_SCN_ATTRIB, Long.toString(ts));

            try (Connection conn = DriverManager.getConnection(getUrl(), props)) {
                conn.createStatement().execute(this.indexDDL);

{code}

My guess is this is some kind of an UPSERT SELECT issue i.e. the scan is seeing the data that
is being inserted which is causing it to run into an infinite loop and timeout. I will try
and get rid of the scn altogether in these tests.

> Tests extending BaseQueryIT are flapping
> ----------------------------------------
>
>                 Key: PHOENIX-4151
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-4151
>             Project: Phoenix
>          Issue Type: Bug
>            Reporter: Samarth Jain
>            Assignee: Samarth Jain
>
> Sample failures:
> {code}
> ERROR] testNotInListOfFloat[NotQueryIT_1](org.apache.phoenix.end2end.NotQueryIT)  Time
elapsed: 0.001 s  <<< ERROR!
> java.lang.RuntimeException: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.NotQueryIT.<init>(NotQueryIT.java:56)
> Caused by: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.NotQueryIT.<init>(NotQueryIT.java:56)
> {code}
> {code}
> [ERROR] testValidStringConcatExpression[indexDDL=CREATE INDEX %s ON %s (a_integer, a_string)
INCLUDE (    B_STRING,     A_DATE) %s,mutable=false,columnEncoded=false](org.apache.phoenix.end2end.QueryIT)
 Time elapsed: 0.014 s  <<< ERROR!
> java.lang.RuntimeException: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.QueryIT.<init>(QueryIT.java:66)
> Caused by: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.QueryIT.<init>(QueryIT.java:66)
> {code}
> {code}
> [ERROR] testNullMultiCondCaseStatement[CaseStatementIT_1](org.apache.phoenix.end2end.CaseStatementIT)
 Time elapsed: 0.007 s  <<< ERROR!
> java.lang.RuntimeException: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.CaseStatementIT.<init>(CaseStatementIT.java:58)
> Caused by: java.sql.SQLTimeoutException: Operation timed out.
> 	at org.apache.phoenix.end2end.CaseStatementIT.<init>(CaseStatementIT.java:58)
> {code}
> My hunch is that timeouts are happening when creating tables or indices. Probably some
sort of a timestamp/scn issue. [~jamestaylor] - any ideas? Would injecting our own clock in
EnvironmentEdge help?



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message