On Mon, Jan 10, 2011 at 7:59 PM, Emmanuel Lecharny <[email protected]> wrote: > Hi, > > it's not the first time I'm annoyed by the way the exception are 'swallowed' > by the server, but now, I think there is something to do to fix that. > > What happens is that we have two kind of exceptions being generated by the > server : > - LDAP related exceptions, those we throw in the server by creating a new > LDAPException. > - other exceptions, like NullPointerException, which are encapsulated in > some way or another > > At the end, from the client side, what we get is an UnexpectedException in > the second case, with no way to know where the exception is coming from. > > Since I started to work on the server almost 6 years ago, I was convinced > that writing StackTraces in the logs was a error, as it could fill the logs > very quickly, and make them grow so fast that any valuable message could be > buried into millions of lines of stack trace. > > Right now, I think that I would rather have those stackTraces instead of > spending hours debugging the server or adding some temporary > e.printStackTraces in some random places. Typically, today, I tracked down a > NPE which was generated by a error message using a null reference to inform > the client about the cause of the error. Isn't it ironical that trying to > give back some info when some error occurred you get a NPE ? No. In fact, it > happens all the time, but we aren't protected against such mistakes. > > Now, I would suggest we add some handling of such exception at the higher > level, ie in core-session. > > Wdyt ?
I share the pain here :)/ one idea is to include the stack trace of the received exception (if it is not of type LdapException) into the new exception being thrown > > -- > Regards, > Cordialement, > Emmanuel Lécharny > www.iktek.com > > -- Kiran Ayyagari
