|
> Ok, this might be ok if you use piranha. But then you have to
> address the problem in the piranha mailinglist. Here we provide
> help for plain vanilla kernel and setup using the provided tools
> from the LVS package.
Actually, I am not sure if it's a piranha problem, or more likely
something that can't be done using LVS. So, I've posted in both lists.
>> [root@cluster root]# ipvsadm -Ln IP Virtual Server version 0.8.1
>> (size=65536) Prot LocalAddress:Port Scheduler Flags ->
>> RemoteAddress:Port Forward Weight ActiveConn InActConn
>> TCP 172.22.48.212:80 lc
>> -> 172.16.0.1:80 Local 80 0 0
>> -> 172.16.0.102:80 Masq 1 0 0
>> -> 172.16.0.101:80 Masq 2 0 0
>
> And that's exactly why I asked you to show me the output. If you read
> your output carefully and reread your first email, you might notice,
> that you wanted to set the director's weight to 8 but you obviously set
> it to 80! And thus your observation is absolutely correct and has to be
> like that. If you want the balancing like you stated RS:RS:LB=1:1:8, you
> have to set it up that way. Currently you run 1:1:80.
Yes sure, this output shows a weight of 80 for the director. But as
piranha continiously updates the weights to reflect the load on each
machine, the weights changed since the first post.
When I tested, it was 1:1:8 for sure. Anyway, the problem was not from
the weights nor lvs, but from my apaches' configuration (see previous
messages).
> I don't follow the piranha mailinglist. I'm glad your second
> problem has been solved too.
It hasn't ! Only the first has been solved, but the second, involving
local nodes and backup director has not been solved yet.
--
Sébastien Bonnet
Ingénieur d'exploitation
Centre de contacts - Experian France
|
