Hi List,
Using 2.2-dev5-c3500c3, I've got both a server and a
servertemplate/server that are marked 'down' due to dns not replying
with (enough) records. That by itself is alright.. (and likely has been
like that for a while so i don't think its a regression.)
But when i perform a 'seemless reload' with a serverstates file it
causes the warnings below for both server and template.:
[WARNING] 095/150909 (74796) : server-state application failed for
server 'x/y', invalid srv_admin_state value '32'
[WARNING] 095/150909 (74796) : server-state application failed for
server 'x2/z3', invalid srv_admin_state value '32'
Is there a way to get rid of these warnings, and if 32 is a invalid
value, how did it get into the state file at all?
## Severely cut down config snippet..:
backend x
server y AppSrv:8084 id 161 check inter 10000 weight 1
resolvers globalresolvers
backend x2
server-template z 3 smtp.company.tld:25 id 167 check inter 10000
weight 10 resolvers globalresolvers
One could argue that my backend x should have a better dns name
configured, if it doesn't exists i apparently messed up something..
For the second x2 backend though isn't it 'normal' to have a template
sizing account for 'future growth' of the cluster? And as such always
have some extra template-servers available that are in 'MAINT /
resolution' state? As such when 2 servers of x2 are up, and the 3rd is
in resolution state, it shouldn't warn on a restart imho as its to be
expected for most setups.?.
Not sure if its a bug or a feature request, but i do think it should be
changed :). Can it be added to some todo list? Thanks.
Thanks and regards,
PiBa-NL (Pieter)