[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [xmlblaster] connection problems

To: xmlblaster at server.xmlblaster.org
Subject: Re: [xmlblaster] connection problems
From: Michael Atighetchi <matighet at bbn.com>
Date: Fri, 25 Apr 2003 21:01:35 -0400
Cc: ccjones at bbn.com
In-reply-to: <3EA9B723.3050809 at marcelruff.info>
References: <20030425220359.GD1197 at bbn.com> <3EA9B723.3050809 at marcelruff.info>
Reply-to: xmlblaster at server.xmlBlaster.org
Sender: owner-xmlblaster at server.xmlBlaster.org
User-agent: Mutt/1.4i
Marcel,

oops - confused XmlBlasterAccess and XmlBlasterConnection. I'll rerun
the test case with XmlBlasterAccess on Monday and let you know. 

Thanks, and have a nice weekend
Michael


On Sat, Apr 26, 2003 at 12:30:59AM +0200, Marcel Ruff wrote:
> Michael Atighetchi wrote:
> 
> >I started looking into the connection problems I reported earlier to
> >the mailing list, and have some further insight now.
> >
> >When I start up the client and cause it to establish a SOCKET
> >connection, I see the following output in the client
> >
> Could you please try your tests with XmlBlasterAccess instead of 
> XmlBlasterConnection
> as we do not support XmlBlasterConnection anymore.
> For example XmlBlaster connection uses the deprecated file recorder with
>    
> /home/matighet/tmp/fileRecorder/tailback-matighetmatighet-to-xmlBlaster.frc
> which is now done by our queue framework (the file recorder will be a native
> queue plugin one day in the future).
> 
> If there are still problems with XmlBlasterAccess we are very willing to 
> fix them,
> please report them with all details to us.
great.


> 
> If the 'Your secretSessionId is invalid ...' still happens with 
> XmlBlasterAccess
> please report it as well.
Ok.

> 
> The NPE in XmlBlasterException i have tried to fix.
Cool.

> 
> Your XPath problem:
> Please send me privately the internal dump of xmlBlaster so i can
> reproduce the same meta key messages and try to reproduce your problem.

Chris Jones (who is working with me at BBN) should have sent out the
trace today.


> 
> thanks
> Marcel
> 
> >
> >initializing new SOCKET connection
> >[Apr 25, 2003 5:48:11 PM INFO  XmlBlasterConnection] Initializing fail save
> >mode
> >[Apr 25, 2003 5:48:12 PM INFO  XmlBlasterConnection-matighet] Loaded 
> >security plugin=htpasswd v
> >ersion=1.0
> >[Apr 25, 2003 5:48:13 PM INFO  XmlBlasterConnection-matighet] Callback 
> >settings: type=RAM onOve
> >rflow=deadMessage onFailure=deadMessage maxEntries=1000 type=SOCKET 
> >oneway=false burstMode.coll
> >ectTime=0
> >[Apr 25, 2003 5:48:13 PM INFO  SocketConnection] Created SOCKET client 
> >connected to 'xxx.xxx.xxx.
> >xxx' on port 7607, callback address is xxx.xxx.xxx.xxx:59341
> >[Apr 25, 2003 5:48:15 PM INFO  SocketCallbackImpl-dummyLoginName] Started 
> >callback receiver plu
> >gin on 'xxx.xxx.xxx.xxx:59341'
> >[Apr 25, 2003 5:48:22 PM INFO  FileRecorder] Found 99 unprocessed tail 
> >back messages in '/home/
> >matighet/tmp/fileRecorder/tailback-matighetmatighet-to-xmlBlaster.frc'.
> >[Apr 25, 2003 5:48:22 PM INFO  FileRecorder] FileRecorder is ready, 
> >max=10000000 tail back mess
> >ages are stored in 
> >'/home/matighet/tmp/fileRecorder/tailback-matighetmatighet-to-xmlBlaster.frc
> >'
> >[Apr 25, 2003 5:48:22 PM INFO  XmlBlasterConnection-matighet] Activated 
> >fail save mode: type=SO
> >CKET oneway=false burstMode.collectTime=0 delay=5000 retries=0 
> >pingInterval=0
> >[Apr 25, 2003 5:48:22 PM INFO  XmlBlasterConnection-matighet] Connected to 
> >xmlBlaster as /node/
> >djmproxy1/client/matighet/-2
> >SOCKET connection connected
> >
> >The client the goes on trying to publish a message.
> >
> >
> >Next, in the Xmlblaster slave node, I see
> >
> >....
> >[Apr 25, 2003 5:47:51 PM [32;40mINFO [0m Main-/node/djmproxy1]  
> >oooooooooooooooooooooooooooooooooooooooooooooo [Apr 25, 2003 5:47:51 PM 
> >[32;40mINFO [0m Main-/node/djmproxy1] o xmlBlaster 0.845+ build 
> >04/25/2003 01:27 PM  o
> >[Apr 25, 2003 5:47:51 PM [32;40mINFO [0m Main-/node/djmproxy1] o READY - 
> >press <?> and <enter> for options    o
> >[Apr 25, 2003 5:47:51 PM [32;40mINFO [0m Main-/node/djmproxy1]  
> >oooooooooooooooooooooooooooooooooooooooooooooo [Apr 25, 2003 5:48:16 PM 
> >[32;40mINFO [0m HandleClientRequest-matighet] Client accepted, coming 
> >from host=/xxx.xxx.xxx.xxx port=59341
> >[Apr 25, 2003 5:48:18 PM [34;40mCALL [0m RequestBroker-/node/djmproxy1] 
> >Entering publish(oid='__sys__UserList', contentMime='text/plain', 
> >contentMimeExtended='null' domain='null' from client 
> >'/node/djmproxy1/client/__RequestBroker_internal[djmproxy1]' ...
> >[Apr 25, 2003 5:48:18 PM [32;40mINFO [0m CorruptionIntruder] 
> >CorruptionIntruder::intercept called
> >[Apr 25, 2003 5:48:18 PM [32;40mINFO [0m PublishRateLimiter] 
> >PublishRateLimiter::intercept called
> >[Apr 25, 2003 5:48:19 PM [32;40mINFO [0m HandleClientRequest-matighet] 
> >Client accepted, coming from host=/xxx.xxx.xxx.xxx port=59344
> >[Apr 25, 2003 5:48:19 PM [33;40mWARN [0m HandleClientRequest-matighet] 
> >Destroying old callback driver 'SOCKEThttp://xxx.xxx.xxx.xxx:7601' ...
> >[Apr 25, 2003 5:48:21 PM [32;40mINFO [0m 
> >DeliveryConnection-callback:/node/djmproxy1/client/matighet/-2] Connection 
> >transition UNDEF -> ALIVE: Success, 
> >callback:/node/djmproxy1/client/matighet/-2 connected.
> >[Apr 25, 2003 5:48:21 PM [34;40mCALL [0m RequestBroker-/node/djmproxy1] 
> >Entering publish(oid='__sys__Login', contentMime='text/plain', 
> >contentMimeExtended='null' domain='null' from client 
> >'/node/djmproxy1/client/__RequestBroker_internal[djmproxy1]' ...
> >[Apr 25, 2003 5:48:21 PM [32;40mINFO [0m CorruptionIntruder] 
> >CorruptionIntruder::intercept called
> >[Apr 25, 2003 5:48:21 PM [32;40mINFO [0m PublishRateLimiter] 
> >PublishRateLimiter::intercept called
> >[Apr 25, 2003 5:48:21 PM [34;40mCALL [0m 
> >ClientSubscriptions-/node/djmproxy1] Login event for client 
> >/node/djmproxy1/client/matighet/-2, nothing to do
> >[Apr 25, 2003 5:48:21 PM [32;40mINFO [0m Authenticate-/node/djmproxy1] 
> >Successful login for client /node/djmproxy1/client/matighet/-2, session 
> >expires after [ 24 h 0 millis ], 1 of 10 sessions are in use.
> >[Apr 25, 2003 5:48:30 PM [32;40mINFO [0m HandleClientRequest-matighet] 
> >Client accepted, coming from host=/xxx.xxx.xxx.xxx port=59360
> >[Apr 25, 2003 5:48:30 PM [33;40mWARN [0m HandleClientRequest-matighet] 
> >Destroying old callback driver 'SOCKEThttp://xxx.xxx.xxx.xxx:7601' ...
> >[Apr 25, 2003 5:48:41 PM [34;40mCALL [0m XmlBlasterImpl-/node/djmproxy1] 
> >Entering publishArr()
> >[Apr 25, 2003 5:48:41 PM [33;40mWARN [0m 
> >Authenticate-/node/djmproxy1.AccessDenied] SessionId '' is invalid, no 
> >access to xmlBlaster.
> >
> >
> >
> >shortly after, I get the following exception on the publishing client.
> >
> >stub trying to publish 
> >errorCode=user.security.authentication.accessDenied message=Your 
> >secretSessionId is invalid, no
> >access to djmproxy1.
> >       at 
> >       org.xmlBlaster.util.XmlBlasterException.parseByteArr(XmlBlasterException.java:560)
> >       at 
> >       org.xmlBlaster.protocol.socket.Parser.getException(Parser.java:475)
> >       at 
> >       org.xmlBlaster.protocol.socket.Executor.receive(Executor.java:357)
> >       at 
> >       org.xmlBlaster.client.protocol.socket.SocketCallbackImpl.run(SocketCallbackImpl.java
> >:165)
> >       at java.lang.Thread.run(Thread.java:536)
> >
> >the code then goes into failover handling, which is trying to reconnect.
> >
> >
> >I also reproduced the same effect with RMI and IOR as transport
> >protocols, which produced a slightly different exception trace.
> >
> >Just to verify that authentication is not the problem I set
> >Security.Server.Plugin.htpasswd.secretfile=NONE
> >in the client's property file, which did not change anything.
> >
> >One more thing: I tried to get more info about the exception using
> >getMessage (as suggest by Marcel earlier), but this only produced a
> >further Nullpointer exception
> >java.lang.NullPointerException
> >       at 
> >       org.xmlBlaster.util.XmlBlasterException.getMessage(XmlBlasterException.java:197)
> >
> >
> >To me it looks as if I'm able to connect to the slave initially, but
> >as soon as a client either publishes or subscribes, xmlblaster has
> >problems with the sessionID and causes the connection to fail.
> >
> >This is based on the xmlblaster cvs tree from today.
> >
> >Any clues ? We are keeping to look into the query string problem and the
> >connection problem next week, but have a hard deadline for the query
> >string functionality this coming wednesday, so try to look at the
> >query string problem first (reported in an earlier email by Chris
> >Jones).
> >
> >One last note: I've attached the slave's property file. Note that I'm
> >running the client, slave, and master on the same machine.
> >
> >Thanks
> >
> >Michael
> >
> >
> >
> >
> >
> >
> > 
> >
> >
> >Security.Server.Plugin.htpasswd.secretfile=NONE
> >
> >CbProtocolPlugin[IOR][1.0]=org.xmlBlaster.protocol.corba.CallbackCorbaDriver
> >CbProtocolPlugin[SOCKET][1.0]=org.xmlBlaster.protocol.socket.CallbackSocketDriver
> >CbProtocolPlugin[RMI][1.0]=org.xmlBlaster.protocol.rmi.CallbackRmiDriver
> >CbProtocolPlugin[XML-RPC][1.0]=org.xmlBlaster.protocol.xmlrpc.CallbackXmlRpcDriver
> >CbProtocolPlugin[JDBC][1.0]=
> >
> >ClientProtocolPlugin[IOR][1.0]=org.xmlBlaster.client.protocol.corba.CorbaConnection
> >ClientProtocolPlugin[SOCKET][1.0]=org.xmlBlaster.client.protocol.socket.SocketConnection
> >ClientProtocolPlugin[RMI][1.0]=org.xmlBlaster.client.protocol.rmi.RmiConnection
> >ClientProtocolPlugin[XML-RPC][1.0]=org.xmlBlaster.client.protocol.xmlrpc.XmlRpcConnection
> >
> >ClientCbServerProtocolPlugin[IOR][1.0]=org.xmlBlaster.client.protocol.corba.CorbaCallbackServer
> >ClientCbServerProtocolPlugin[SOCKET][1.0]=org.xmlBlaster.client.protocol.socket.SocketCallbackImpl
> >ClientCbServerProtocolPlugin[RMI][1.0]=org.xmlBlaster.client.protocol.rmi.RmiCallbackServer
> >ClientCbServerProtocolPlugin[XML-RPC][1.0]=org.xmlBlaster.client.protocol.xmlrpc.XmlRpcCallbackServer
> >
> >
> >jacorb.verbosity=0
> >ns=false
> >port=7601
> >
> >useTopicStore=false
> >JdbcDriver.drivers=
> >
> ># turn off telnet access
> >admin.remoteconsole.port=-1
> >
> ># turn off JDBC plugin
> >ProtocolPlugin[JDBC][1.0]=
> >
> >
> ># clustering
> >cluster=true
> >cluster.node.id=djmproxy1
> >
> >cluster.node[psq1]=\
> >  <clusternode id='psq1'>\
> >     <info>\
> >        <address type='IOR' hostname='xxx.xxx.xxx.xxx' port='7901' />\
> >     </info>\
> >     <master type='DomainToMaster'>\
> >        <key queryType='DOMAIN' domain='*'/>\
> >     </master>\
> >  </clusternode>
> >
> ># publish length checking
> >## only let through first image
> >MimePublishPlugin[PublishRateLimiter][1.0]=com.bbn.quo.dpasa.PublishRateLimiter,MAX_RATE_MPH=3
> >MimePublishPlugin[PublishLenChecker][1.0]=com.bbn.quo.dpasa.PublishLenChecker,DEFAULT_MAX_LEN=46000
> >MimePublishPlugin[CorruptionIntruder][1.0]=com.bbn.quo.dpasa.CorruptionIntruder
> >
> >
> > 
> >
> 
> 

-- 
matighet at bbn.com   BBN Technologies
Follow-Ups:
- Re: [xmlblaster] connection problems
  - From: Michael Atighetchi <matighet at bbn.com>
References:
- [xmlblaster] connection problems
  - From: Michael Atighetchi <matighet at bbn.com>
- Re: [xmlblaster] connection problems
  - From: Marcel Ruff <mr at marcelruff.info>
Prev by Date: [xmlblaster] [xmlblaster-devel] ANNOUNCE xmlBlaster release 0.846
Next by Date: [xmlblaster] Dravis Group OSS report with xmlBlaster
Previous by thread: Re: [xmlblaster] connection problems
Next by thread: Re: [xmlblaster] connection problems
Index(es):
- Date
- Thread