[
https://issues.apache.org/jira/browse/AMQCPP-422?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mikhail Melamud updated AMQCPP-422:
-----------------------------------
Description: When we specify self-signed certificate in PEM-formatted file
as value of the "decaf.net.ssl.trustStore" property and SAN that does't match
any name in failover URI, cms is crashing on "R6025 Pure Virtual Function
Call" run-time error. I think, the issue can be attributed to multithreading...
After ??Server Certificate Name doesn't match the URI Host Name value.??
exception is thrown, CMS is trying to fire an exception using listener pointer
that is not instantiated properly in _void TransportFilter::fire( const
decaf::lang::Exception& ex )_ method... It is not happening all the time, so
a simple unit test won't do... I'm attaching a bit re-factored
_HelloWorldProducer_ app and the certificate file that I used... There is
possibility that it has something to do with InactivityMonitor injected into
transport filter chain... When I tried
{color:blue}"failover://(ssl://dev1467:61618?transport.useInactivityMonitor=false){color}
instead of {color:blue}"failover://(ssl://dev1467:61618)"{color} it behaves
much more stable, i.e. client app can be re-run many more times without a
crash, still crashing though in OpenFormatNegotiator (was: When we specify
self-signed certificate in PEM-formatted file as value of the
"decaf.net.ssl.trustStore" property and SAN that does't match any name in
failover URI, cms is crashing on "R6025 Pure Virtual Function Call" run-time
error. I think, the issue can be attributed to multithreading... After ??Server
Certificate Name doesn't match the URI Host Name value.?? exception is thrown,
CMS is trying to fire an exception using listener pointer that is not
instantiated properly in _void TransportFilter::fire( const
decaf::lang::Exception& ex )_ method... It is not happening all the time, so
a simple unit test won't do... I'm attaching a bit re-factored
_HelloWorldProducer_ app and the certificate file that I used... There is
possibility that it has something to with InactivityMonitor injected into
transport filter chain... When I tried
{color:blue}"failover://(ssl://dev1467:61618?transport.useInactivityMonitor=false){color}
instead of {color:blue}"failover://(ssl://dev1467:61618)"{color} it behaves
much more stable, i.e. client app can be re-run many more times without a
crash, still crashing though in OpenFormatNegotiator)
> When we specify self-signed certificate in PEM-formatted file as value of the
> "decaf.net.ssl.trustStore" property and SAN that does't match any name in
> failover URI, cms is crashing on "R6025 Pure Virtual Function Call" run-time
> error.
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: AMQCPP-422
> URL: https://issues.apache.org/jira/browse/AMQCPP-422
> Project: ActiveMQ C++ Client
> Issue Type: Bug
> Components: CMS Impl, Decaf
> Affects Versions: 3.4.0, 3.4.2, 3.4.4
> Environment: Broker: 5.6.0,
> CMS 3.4.2
> Windows
> Reporter: Mikhail Melamud
> Assignee: Timothy Bish
> Attachments: AMQTest.cpp, brokerkeystore.pem
>
>
> When we specify self-signed certificate in PEM-formatted file as value of the
> "decaf.net.ssl.trustStore" property and SAN that does't match any name in
> failover URI, cms is crashing on "R6025 Pure Virtual Function Call" run-time
> error. I think, the issue can be attributed to multithreading... After
> ??Server Certificate Name doesn't match the URI Host Name value.?? exception
> is thrown, CMS is trying to fire an exception using listener pointer that is
> not instantiated properly in _void TransportFilter::fire( const
> decaf::lang::Exception& ex )_ method... It is not happening all the time,
> so a simple unit test won't do... I'm attaching a bit re-factored
> _HelloWorldProducer_ app and the certificate file that I used... There is
> possibility that it has something to do with InactivityMonitor injected into
> transport filter chain... When I tried
> {color:blue}"failover://(ssl://dev1467:61618?transport.useInactivityMonitor=false){color}
> instead of {color:blue}"failover://(ssl://dev1467:61618)"{color} it behaves
> much more stable, i.e. client app can be re-run many more times without a
> crash, still crashing though in OpenFormatNegotiator
--
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
