high availability with 2 gnugks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



hi list 

[quite lenghty post]

i'm trying to set up a high availability GK service with two gnugks. 

for that, it seems there are two solutions that don't involve virtual servers mechanisms: (setup1) using the alternate gatekeepers mechanism, or (setup2) configuring "mutual" neighbors. there are caveats for both setups, and i'm wondering if some of you had similar setups in production. 


the setup is the following:

- GK1 and GK2 are the "main" gks, and are physically located at 2 different places
the endpoints that will register are very heterogeneous (soft/hard terminals, gateways, gks, ...); 
- some endpoints are with real ips, and some are behind hosts that do nat. the endpoints are located all over europe, so i can't configure GK1 or GK2 to support nated endpoints ( eg. routing the rtp traffic for a france-france call via sofia wouldn't be the best thing to do ). ; instead, i use gnugk configured as proxy on each NAT host.


setup1, with GK1 and GK2 using the alternategks= config :

endpoints usually provide an "alternate gk" configuration option; enpoints with real ips are configured with primary=gk1, and secondary=gk2

each gnugk on NAT host registers as an endpoint to GK1, with UseAlternateGK=1 ; Gk1 will send "GK2" as alternate id in the RCF, and the gnugk endpoint should also register with GK2 (no tested though); that's nice, but what if GK1 is down when the endpoint tries to register ? i've tried - without success - to set up two [endpoints] sections to register to GK1 and GK2, or with 2 gatekeeper= statements. (feature request :) )

however, some posts suggest not to use the alternategk mechanism ; but except if GK1 and/or GK2 are also configured as proxies for a physically linked LAN, i don't see what can fail - considering that endpoints are not buggy. any precision ? maybe that would be nice to add this to the manual.


setup2, with GK1 and GK2 as neighbors of each other.

as above, endpoints with real ips that support alternate gks use prim=gk1 and sec=gk2.

GK1,GK2, and gngks on NAT hosts are now configured as neighbors, eg: gk1 and gk2 sends prefix XX to gknat_1, and gknat_1 sends prefixes "*" to gk1 with prio 1, and "*" to gk2 with prio 2.

caveat: increased complexity and timeouts: NeighborTimeout timeout when gknat_X is down (there would be no timeout in setup1 if gknat_X was not registered).



any information/suggestion/precision will be welcomed !

thanks
ivan


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/

_______________________________________________________

List: Openh323gk-users@xxxxxxxxxxxxxxxxxxxxx
Archive: http://sourceforge.net/mailarchive/forum.php?forum_id=8549
Homepage: http://www.gnugk.org/

[Index of Archives]     [SIP]     [Open H.323]     [Gnu Gatekeeper]     [Asterisk PBX]     [ISDN Cause Codes]     [Yosemite News]

  Powered by Linux