Thanks for your help,
what do you mean with take it off the server farms first?
The upgrade process is like this:
stop tomcat
change symlink to new version
start tomcat
loadbalancer detect that the server is down and routes the traffic
to the other hosts.
This is like we always do since tomcat 6.
After downgrading back to the old version load is comming back fast.
Because of the shared infrastructure every change to the
loadbalancing pools must be ordered and costs money.
On 07/31/2017 07:04 PM, M. Manna wrote:
When you upgraded the affected tomcat from 8.0.45 to 8.5.16, Did you take
it off the server farms first? Or did you do it without?
Try to remove the affected tomcat off the server farm. Do you upgrade, and
then put it back inside the farm.
Or have you tried this already?
On 31 July 2017 at 17:59, Bernd Wahlen <bernd.wah...@k2interactive.de>
wrote:
We are running a cluster of 5 tomcats behind
a loadbalancer (shared big ip f5).
Traffic is http, https and websocket.
After upgrading one of the servers to
Tomcat 8.5.16 this server get no traffic anymore.
Everything is running fine until 8.0.45.
If i connect directly against the node
with 8.5, everything is running fine also,
so it must be a issue related to the loadbalancing
or loadbalancing checks.
I read the changes and tries the following
1.)
https://www.joedog.org/2016/02/28/test-connectivity-to-an-aj
p-server-with-ajping/
=> AJP is still working
2.)
Configure CookieProcessor to LegacyCookieProcessor
but no success.
Any idea what to i can check or what changed related
to that problem?
Many thanks in advance,
Bernd Wahlen
K2 Interactive GmbH
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
