Memcache Ticket Registry HA

[Teddy Brown]

Is it possible to get High Availability with the memcache ticket registry? 

I only have these attributes configured currently and it works.

However it seems if the Memcached instance on either host is restarted (or the host is restarted) that CAS continues to function as expected, any tickets on the restarted host need to re-authenticate.
cas.ticket.registry.memcached.servers=cas01:11211,cas02:11211
cas.monitor.memcached.failureMode=Redistribute

Is it possible to configure this in a way that a restart of either Memcached service will not result in the loss of any CAS tickets? 

Thanks

[Ray Bon]

Teddy,

I have not used memcached. To accomplish your goal you would need a replicated cache.

How often do you plan to restart your servers? Will your users to notice?

Ray

-- 
Ray Bon
Programmer analyst
Development Services, University Systems
2507218831 | CLE 019 | rb...@uvic.ca

[Julien Gribonvald]

Hi,

You can use repcached for the replicated side with memcached, we are using it since several years in our context and we are totally satisfied !

Thanks

--
- Website: https://apereo.github.io/cas
- Gitter Chatroom: https://gitter.im/apereo/cas
- List Guidelines: https://goo.gl/1VRrw7
- Contributions: https://goo.gl/mh7qDG
---
You received this message because you are subscribed to the Google Groups "CAS Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cas-user+u...@apereo.org.
To view this discussion on the web visit https://groups.google.com/a/apereo.org/d/msgid/cas-user/1523375473.1822.5.camel%40uvic.ca.

--
Julien Gribonvald

[Teddy Brown]

That looks very promising!  We've been using memcached for some time as a cache but this is the first time I've ever looked for a replicated model.  Thank you!

[Teddy Brown]

We perform regular monthly server updates, occasionally more as needed.  If we had ticket replication (which Memcached itself doesn't support) we could do these updates during regular working hours and users wouldn't notice. 

Doing maintenance off hours is an option, and how we do it for services which don't have high availability.  Users needing to re-authenticate is acceptable in the event of a server failure as long as CAS remained online, but we would not perform regular maintenance during on hours if it may event in some users being forced to re-reauthenticate.