[jcifs] Problem with FCIFS-1.2.6 - cry for assistance

MIchael B. michael.bunger at amdocs.com
Tue Oct 3 22:34:04 GMT 2006

Hellow; My name is Michael; I was so glad to find such an informative forum

Can you please look at the next problem? May be U'll have a sugestion how to 
resolve it

We are using jcifs-1.2.6 / WLS 8.1 SP 4 + the following the part of 


This is what we get in the log: (It's not happen a lot, but when happens the 
first time requerd server restart -> affects static Indx value in NbtAddress

public static InetAddress getWINSAddress() { 
        return NBNS.length == 0 ? null : NBNS[nbnsIndex]; 


<Oct 2, 2006 8:51:55 PM EDT> <Error> <HTTP> <BEA-101020> <[ServletContext
(id=18073170,name=webApp_crmServer,context-path=)] Servlet 
failed with Exception
java.lang.ArrayIndexOutOfBoundsException: 2
        at jcifs.netbios.NbtAddress.getWINSAddress(NbtAddress.java:524)
        at jcifs.UniAddress.getByName(UniAddress.java:267)
        at jcifs.http.NtlmHttpFilter.negotiate(NtlmHttpFilter.java:156)
        at jcifs.http.NtlmHttpFilter.doFilter(NtlmHttpFilter.java:114)
        at weblogic.servlet.internal.FilterChainImpl.doFilter
        at weblogic.security.acl.internal.AuthenticatedSubject.doAs
        at weblogic.security.service.SecurityManager.runAs
        at weblogic.servlet.internal.WebAppServletContext.invokeServlet
        at weblogic.servlet.internal.ServletRequestImpl.execute
        at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:219)
        at weblogic.kernel.ExecuteThread.run(ExecuteThread.java:178)

We found this in 

It was discovered that an ArrayIndexOutOfBoundsException could occur if the 
list  of domain controllers returned by NbtAddress.getAllByName was shorter 
than  the  list  returned  in  the previous call (possibly because the WINS 
query  timed  out  and switched to an alternate WINS server). All NTLM HTTP 
Authentication Filter users should upgrade to prevent this error. Also, the 
value  of  jcifs.netbios.cachePolicy  set  by the NTLM HTTP Filter if it is 
not specified has been doubled to 20 minutes. Finally, some log levels have 
been  increased  such that running with jcifs.util.loglevel = 3 temporarily 
is  actually  reasonable in a production environment (must use loglevel > 3 
to see individual SMB messages and loglevel > 5 to get hexdumps).  

I'm wandering to know, if U see this problem in the past & if 1.2.9 upgrade can 
fix it

Thank U so much in advance

More information about the jcifs mailing list