-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I managed to recreate this by waiting and killing the node that authed a
user immediately after the test case ran. I have had a hard time doing
this by hand as I can't seem to kill it fast enough...

Jeff


Jeff Chapin wrote:
> I have CAS clustered, and mostly working.
> 
> I have TGT stored in the database (but at this time the ST are not -- I
> opened a second email over this issue), and I am using the
> repository-type="client" method.
> 
> If you log in, and I shut down the node that you logged in to,
> refreshing the login page still results in a success.
> 
> When either one node, or both nodes are up, no issues occur. Our test
> users can log in, access a protected service, access a second, and then
> log out.
> 
> If one of the two nodes shutsdown, occasionally reach a screen saying
> you need to relogin -- but I am not able to recreate this outside of our
> test script.
> 
> This happens even if I explicitly force the cluster name into the cookie.
> 
> Any ideas what is going on? I am trying to figure out what *exactly*
> causes this, but I figured someone might know.
> 
> Jeff
> 

- --
Jeff Chapin,
Assistant Systems/Applications Administrator
ITS-IS, University of Northern Iowa
Phone: 319-273-3162 Email: [email protected]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAkvbJwMACgkQQiaEUfQoY7SCJwCgxrhjTWcrcjbsubEDaZ4LyiZG
4f4AoL4z2a60Ktga0zSdHMu699AC9OXt
=BSEe
-----END PGP SIGNATURE-----

-- 
You are currently subscribed to [email protected] as: 
[email protected]
To unsubscribe, change settings or access archives, see 
http://www.ja-sig.org/wiki/display/JSG/cas-user

Reply via email to