On Fri, Sep 9, 2011 at 8:34 PM, Yuji Shinozaki <span dir="ltr"><<a href="mailto:ys2n@virginia.edu">ys2n@virginia.edu</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im"><br>
> <LDAPProperty name="edu.vt.middleware.ldap.operationRetryExceptions" value="javax.naming.CommunicationException,javax.naming.ServiceUnavailableException,javax.naming.TimeLimitExceededException" /><br>
><br>
<br>
</div>Thanks. That makes more sense. That syntax is unclear in the documentation.<br>
<div class="im"><br>
> Are you sure this is what you want? An operation retry closes and reopens the connection, then presumably you'd get the same TimeLimitExceededException again. Note that LimitExceededException is ignored by the search result handler. So even if it occurs, you'll still get any results that were retrieved before the exception.<br>
><br>
<br>
</div>Ok. This may be a matter of upgrading our shib implementation (it is getting quite old) as the error is resulting in an attribute resolution error bubbling all the way up. We are getting the time limit exceeded exceptions at irregular intervals, but with increasing frequency lately. So i was hoping a properly-spaced retry would workaround these brown-outs. Our campus ldap admins have reassured us that they are throwing new hardware at the general problem, but of course that won't be for the proverbial "few weeks".<br>
<br></blockquote><div><br></div><div>Since you have a misbehaving LDAP your approach may help triage the issue. Assuming these time limit exceptions are load related, you should also play around with a retry backoff. Stepping down the load might be your best bet.</div>
<div><br></div><div>--Daniel Fisher</div><div><br></div></div>