Re: udp port support
pb <[email protected]>
| Newsgroups | gmane.linux.redhat.piranha |
|---|---|
| Message-ID | <[email protected]> |
Hello Sebastien, Yes, that worked - I added a external script which echos "OK" to match the expected result "OK", and now my routes stay put as desired. In this case, I have several VIP's for direct administrative access (via LVS) to each of the Real Servers, so those don't really need to lose their routes when those boxes go down. On the other hand the one VIP is set aside for a publicly accessible LOAD BALANCING LDAP SERVICE which does need its external script to actually work and test for those boxes to be live or not as far as LDAP ports. Thanks again, Peter --- Sébastien Bonnet <[email protected]> wrote: > > I am trying to support snmp and snmptrap [...] > > but what happens is, the PULSE daemon takes a few > > seconds to load LVS and NANNY daemons, so this > > NAT routing for those UDP ports gets dropped > > soon as LVS loads NANNY daemons. > > Are you sure this is what happen ? I'd say nannies > are started and > because you don't have the approriate send/expect > string, nannies remove > the RS. > > Try an external script (send_program) which just > echoes OK instead of > the send string and set the expect string to OK. > This should solve the > problem if it is what I think. You'll then have to > rewrite this script > to really ensure that the SNMP stuff *is* working OK > on the RS. > > Regards, > > -- > Sébastien Bonnet > __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com