Shibboleth SP clustering using shared database
wlee007-m at yahoo.com
Tue Feb 20 14:47:50 EST 2018
Besides RAC, we also tried memcached as StorageService, similar results: when one of memcached is offline, SP would just error out, doesn't seem to skip the bad one and try the other memcached. Since RAC happened to be available to us, we tried it with the hope that ODBC driver would fill the gap of failover but it didn't.
On Thursday, February 15, 2018, 10:39:26 AM EST, Cantor, Scott <cantor.2 at osu.edu> wrote:
> We have tried with Oracle RAC database cluster, but no luck:
> 1. When one of Oracle RAC node is offline, SP stops working due to unable to
> connect to database (StorageService), it turned out Oracle ODBC driver does
> not support transparent fail over, their JDBC driver does.
That's kind of hysterical given how much RAC costs. Sort of defeats the entire point, doesn't it?
Given where the client-side storage option is likely headed wrt timeouts, given the way apps have to be built to essentially ignore timeouts, and given that timeouts are just plain stupid for the most part, I think it would be within the realm of sanity to enhance the cache to optionally ignore errors during last-access touch attempts. That's assuming it doesn't already, but I don't recall.
For Consortium Member technical support, see https://wiki.shibboleth.net/confluence/x/coFAAg
To unsubscribe from this list send an email to users-unsubscribe at shibboleth.net
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the users