On 1/4/21 3:11 AM, Eric Robinson wrote:
-----Original Message----- From: Users
<[email protected]> On Behalf Of Timo Schöler Sent:
Saturday, January 2, 2021 4:27 AM To: [email protected]
Subject: Re: [ClusterLabs] Observed Difference Between ldirectord
and keepalived
On 1/2/21 10:11 AM, Eric Robinson wrote:
We recently switched from ldirectord to keepalived. We noticed
that, after the switch, LVS behaves a bit differently with
respect to "down" services.
On ldirectord, a virtual service with 2 realservers displays
"Masq 0" when one of them is down.
TCP 192.168.5.100:3002 wlc persistent 50
-> 192.168.8.53:3002 Masq 1 0 0
-> 192.168.8.55:3002 Masq 0 0 4
On keepalived, it does not shown the down server at all...
TCP 192.168.5.100:3002 wlc persistent 50
-> 192.168.8.53:3002 Masq 1 0 0
Why is that? It makes it impossible to see when services are
down.
Hi Eric,
I think you're showing the output of ``ipvsadm -Ln'' here, right?
If so, what OS are you using? I observed similar behaviour when
moving from CentOS 6 to CentOS 7 (i.e. Kernel 2.6.x to Kernel
3.10.x), so there may have changed default values of displaying
stuff or APIs.
Timo
Hi Timo,
That is exactly the case! The old load balancer is on CentOS 6 and
the new one is on CentOS 7. (I'm aware of the recent announcement
regarding CentOS, but 7 will not be end-of-life for another 4
years.)
Did you ever come up with a solution? We have scripts that rely on
the old behavior to alert admins of down services.
-Eric
Hi Eric,
world's moving on (regarding CentOS)...
I had the same issue, icinga2 was in the game. I modified the custom
made NRPE checks making them differentiate between OS versions.
Timo
_______________________________________________
Manage your subscription:
https://lists.clusterlabs.org/mailman/listinfo/users
ClusterLabs home: https://www.clusterlabs.org/
