ldap queries hanging

[Daniel Fisher]

On Fri, Dec 16, 2016 at 10:04 AM, Ted Wood <TWood at parker.edu> wrote:

> I already have this rule in the connectionConfig bean:
>
>
>
> p:connectTimeout="%{idp.authn.LDAP.connectTimeout:3000}"
>
>
>
> On its face it seems like it would do the same thing but is there a
> difference?
>

There is. One controls timeout for new connections, the other controls
timeouts for LDAP operation reads.

  I’ve also noticed a bunch of these messages in the idp-warn.log:
>
>
>
> WARN [org.ldaptive.AbstractOperation$ReopenOperationExceptionHandler:277]
> - Operation exception encountered, reopening connection
>

These messages can indicate that your LDAP or load balancer is dropping
connections.
Tuning your pool config by lowering the validatePeriod may keep your
connections alive.



> I found a reference to this warning here:
>
>
>
> http://shibboleth.net/pipermail/users/2015-March/020576.html
>
>
>
> and this is supposed to be a bug that was fixed in ldaptive 1.0.6 which I
> am already using.  Perhaps this is an edge case and would there be an
> problem with upgrading just the ldaptive jar to 1.2.1 which appears to be
> the latest version?
>

Unless you're seeing that *exact* stacktrace, then this bug is unrelated.
ldaptive v1.2.1 will not work with the current IDP.

--Daniel Fisher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://shibboleth.net/pipermail/users/attachments/20161216/50029362/attachment-0001.html>