Hi dear support,
While migrating a critical RFA 7.6.1_L1 C++ application in production from SSLED to RSSL adapter (connected to ADS 2.6.2_L1 linux), the RFA crashes at startup with the following error message :
--------
RFA Internal failure
Application: RFA
Process Id: 0x5788
File: Session_OMM/Impl/OMMSrcDirectoryMsgHandler.cpp
Line: 1039
Problem:
RFA_VERIFY( rsslDecodeMsg( &dec_mapIter, &newRsslMsg ) >= RSSL_RET_SUCCESS ) failed
---------
After some investigation, it appears that we are able to reproduce the issue only when connecting to an ADS with a consequent number of services (126 exactly).
This is the stack trace of the process when it crashes :
----------
(gdb) bt
#0 0xf77a6430 in __kernel_vsyscall ()
#1 0xf1a2c861 in raise () from /lib/libc.so.6
#2 0xf1a2e0ca in abort () from /lib/libc.so.6
#3 0x08a4923e in __RFA_ProblemReport ()
#4 0x08a29b6d in rfa::sessionLayer::OMMSrcDirectoryMsgHandler::convertServiceDirectoryMapKeyFromSrcIdToServiceName(rfa::sessionLayer::OMMItemEventImplMsg&, RwfBuffer&) ()
#5 0x08a29cc9 in rfa::sessionLayer::OMMSrcDirectoryMsgHandler::notifyClientServiceDirectoryMsg(rfa::sessionLayer::OMMItemEventImplMsg const*) ()
#6 0x08a2cda2 in rfa::sessionLayer::OMMSrcDirectoryMsgHandler::processRefreshUpdateMsg(rfa::sessionLayer::OMMItemEventImplMsg const*) ()
#7 0x08a2ce41 in rfa::sessionLayer::OMMSrcDirectoryMsgHandler::processResponseMsg(rfa::sessionLayer::OMMItemEventImplMsg*, rfa::common::HandleInt*, rfa::sessionLayer::OMMWatchlistEntry*) ()
#8 0x08a03155 in rfa::sessionLayer::OMMMsgRouter::processResponseMsg(rfa::support::SmartPtr<rfa::common::Message> const&) ()
#9 0x089f8f7a in rfa::sessionLayer::OMMConsumerImpl::processResponseMsg(rfa::support::SmartPtr<rfa::common::Message>&) ()
#10 0x089d60d6 in rfa::sessionLayer::SessionImpl::dispatchResponseProcedure() ()
#11 0x089d4275 in rfa::sessionLayer::SessionImpl_ResponseProcessingTimer::processTimerEvent() ()
#12 0x08a6a94c in RTREventNotifierImp::expireEvents() ()
#13 0x08a69925 in RTRSelectNotifier::enable() ()
#14 0x089d69e2 in rfa::sessionLayer::SessionImpl::runThread() ()
#15 0x08a496fa in rfa::support::Thread::startWrapper(void*) ()
#16 0xf2e14bc9 in start_thread () from /lib/libpthread.so.0
#17 0xf1ae4dee in clone () from /lib/libc.so.6
--------------------
I attached a RFA log + trace files, where I believe it seems to show that the failure happens while the RFA requests to download the DataDictionary from the ADS.
Also, I dumped the RIPC session from the ADS, showing that after sending the "large" services map to this client, no more event comes from the client RFA (this is the latest message received or sent by the ADS to this client) :
---------
[Mon Oct 17 17:20:15.034340 2016] RSSL Message Of Length 28 Received on Socket 19
<!-- rwfMajorVer="14" rwfMinorVer="1" -->
<requestMsg domainType="RSSL_DMT_DICTIONARY" streamId="3" containerType="RSSL_DT_NO_DATA" flags="0x40 (RSSL_RQMF_HAS_QOS)" qosDynamic="0" qosRate="1" qosTimeliness="1" dataS ize="0">
<key flags="0xB (RSSL_MKF_HAS_SERVICE_ID|RSSL_MKF_HAS_NAME|RSSL_MKF_HAS_FILTER)" serviceId="9365" name="RWFFld" filter="7"/>
<dataBody>
</dataBody>
</requestMsg>
------------
Could you try to explain what happens here exactly, and if there is a possibility to avoid this ?
Thanks in advance for your help.
Kind regards,
Julien
