You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@activemq.apache.org by "Mark (JIRA)" <ji...@apache.org> on 2012/05/16 00:59:06 UTC
[jira] [Created] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Mark created AMQ-3842:
-------------------------
Summary: test units: disk space requirement excessive/not checked transparently enough
Key: AMQ-3842
URL: https://issues.apache.org/jira/browse/AMQ-3842
Project: ActiveMQ
Issue Type: Bug
Components: Test Cases
Affects Versions: 5.6.0
Environment: any
Reporter: Mark
Priority: Trivial
persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hangs.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13279187#comment-13279187 ]
Mark commented on AMQ-3842:
---------------------------
You are correct. Changed topic and description.
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mark updated AMQ-3842:
----------------------
Description:
persistent DB requires at least 50.000 MB free temp. disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.
was:
persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.
> test units: disk space requirement excessive/not checked transparently enough
> -----------------------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: any
> Reporter: Mark
> Priority: Trivial
>
> persistent DB requires at least 50.000 MB free temp. disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
> 1.) Do we really need 50 GB free disk space for the tests?
> 2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Closed] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Timothy Bish (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Timothy Bish closed AMQ-3842.
-----------------------------
Resolution: Not A Problem
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
> My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
> The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units (and also produces spam packets on the local network during testing :-)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mark updated AMQ-3842:
----------------------
Description:
Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units.
was:Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
> My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
> The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Timothy Bish (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13279248#comment-13279248 ]
Timothy Bish commented on AMQ-3842:
-----------------------------------
The tests aren't really designed to be portable, they are designed to test the various use cases of the Broker. You may need to make accommodations on your specific machine to allow all tests to run. I've added autofail to the test case you mentioned which should prevent the hang, it will still fail of course.
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
> My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
> The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units (and also produces spam packets on the local network during testing :-)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mark updated AMQ-3842:
----------------------
Description:
persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.
was:
persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hangs.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.
> test units: disk space requirement excessive/not checked transparently enough
> -----------------------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: any
> Reporter: Mark
> Priority: Trivial
>
> persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
> 1.) Do we really need 50 GB free disk space for the tests?
> 2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mark updated AMQ-3842:
----------------------
Regression: Unit Test Broken
Description: Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units. (was: persistent DB requires at least 50.000 MB free temp. disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
1.) Do we really need 50 GB free disk space for the tests?
2.) The error should trigger an immediate test failure.)
Environment: NAT FW (was: any)
Labels: discovery (was: )
Summary: test units should run discovery mechanism via loopback device (was: test units: disk space requirement excessive/not checked transparently enough)
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (AMQ-3842) test units should run discovery
mechanism via loopback device
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mark updated AMQ-3842:
----------------------
Description:
Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units (and also produces spam packets on the local network during testing :-)
was:
Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units.
> test units should run discovery mechanism via loopback device
> -------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: NAT FW
> Reporter: Mark
> Priority: Trivial
> Labels: discovery
>
> Many test units are failing on boxes with NAT firewalls enabled, although discovery via loopback device would (IMHO) be sufficient for test units.
> My NAT setup allows all traffic via lo, but only related traffic via eth0 etc., which should be the regular fw security setup on most or many end-user boxes.
> The test org.apache.activemq.transport.discovery.DiscoveryTransportNoBrokerTest (for example) makes activemq send out discovery messages via eth0, which is probably not a perfect solution if one wants to have portable test units (and also produces spam packets on the local network during testing :-)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Posted by "Timothy Bish (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13279132#comment-13279132 ]
Timothy Bish commented on AMQ-3842:
-----------------------------------
The error messages simply point out that the configuration could lead to issues. The tests themselves don't really need 50gig. Why its hanging probably doesn't have any relation to the error especially for the test case shown. You could easily validate this by changing the test to reduce the configured min space requirement and see if it still hangs.
> test units: disk space requirement excessive/not checked transparently enough
> -----------------------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: any
> Reporter: Mark
> Priority: Trivial
>
> persistent DB requires at least 50.000 MB free temp. disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
> 1.) Do we really need 50 GB free disk space for the tests?
> 2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Posted by "Timothy Bish (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13277775#comment-13277775 ]
Timothy Bish commented on AMQ-3842:
-----------------------------------
Which tests are actually hanging?
> test units: disk space requirement excessive/not checked transparently enough
> -----------------------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: any
> Reporter: Mark
> Priority: Trivial
>
> persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
> 1.) Do we really need 50 GB free disk space for the tests?
> 2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (AMQ-3842) test units: disk space requirement
excessive/not checked transparently enough
Posted by "Mark (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/AMQ-3842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13278015#comment-13278015 ]
Mark commented on AMQ-3842:
---------------------------
current trunk:
mvn test -V -B -U -Dtest=DiscoveryTransportNoBrokerTest
tail -F /home/mark/tmp/activemq/activemq-core/target/activemq-test.log :
2012-05-17 19:46:20,350 [JMX connector ] - INFO ManagementContext - JMX consoles can connect to service:jmx:rmi:///jndi/rmi://localhost:1099/jmxrmi
2012-05-17 19:46:20,380 [main ] - INFO PListStore - PListStore:[/home/mark/tmp/activemq/activemq-core/target/activemq-data/localhost/tmp_storage ] started
2012-05-17 19:46:20,382 [main ] - INFO BrokerService - Using Persistence Adapter: KahaDBPersistenceAdapter[/home/mark/tmp/activemq/activemq-core/target/activemq-data/localhost/KahaDB]
2012-05-17 19:46:20,708 [main ] - INFO MessageDatabase - KahaDB is version 4
2012-05-17 19:46:20,726 [main ] - INFO MessageDatabase - Recovering from the journal ...
2012-05-17 19:46:20,726 [main ] - INFO MessageDatabase - Recovery replayed 1 operations from the journal in 0.008 seconds.
2012-05-17 19:46:20,733 [main ] - INFO BrokerService - ActiveMQ null JMS Message Broker (localhost) is starting
2012-05-17 19:46:20,733 [main ] - INFO BrokerService - For help or more information please see: http://activemq.apache.org/
2012-05-17 19:46:20,878 [main ] - INFO TransportServerThreadSupport - Listening for connections at: tcp://localhost:36299?transport.closeAsync=false
2012-05-17 19:46:20,881 [main ] - INFO TransportConnector - Connector tcp://localhost:36299?transport.closeAsync=false Started
2012-05-17 19:46:20,882 [main ] - INFO BrokerService - ActiveMQ JMS Message Broker (localhost, ID:i5-34740-1337276780746-0:1) started
2012-05-17 19:46:20,883 [main ] - WARN BrokerService - Store limit is 100000 mb, whilst the data directory: /home/mark/tmp/activemq/activemq-core/target/activemq-data/localhost/KahaDB only has 39120 mb of usable space
2012-05-17 19:46:20,883 [main ] - ERROR BrokerService - Temporary Store limit is 50000 mb, whilst the temporary data directory: /home/mark/tmp/activemq/activemq-core/target/activemq-data/localhost/tmp_storage only has 39120 mb of usable space
2012-05-17 19:46:20,883 [main ] - INFO DiscoveryTransportNoBrokerTest - Broker started - thread Count:8
2012-05-17 19:46:20,883 [main ] - INFO DiscoveryTransportNoBrokerTest - Connecting.
> test units: disk space requirement excessive/not checked transparently enough
> -----------------------------------------------------------------------------
>
> Key: AMQ-3842
> URL: https://issues.apache.org/jira/browse/AMQ-3842
> Project: ActiveMQ
> Issue Type: Bug
> Components: Test Cases
> Affects Versions: 5.6.0
> Environment: any
> Reporter: Mark
> Priority: Trivial
>
> persistent DB requires at least 50.000 MB free disk space at broker startup during tests. If there is less available, an error is logged to active-mq.log, but the tests just hang.
> 1.) Do we really need 50 GB free disk space for the tests?
> 2.) The error should trigger an immediate test failure.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira