The customer has the following problem. He uses cookie insertion by ACE4710 (ver A3(2.0)). All is working fine, but there is one exception:-( When the client uses VIP, the ACE load-balanced this request and "sticky" the client into server s1 (for example). The client has cookie for s1. But when this server s1 fails, the client will get no web pages - it seems like that ACE has a problem with cookie for the server which is failed. I saw the increasing count of failed connection in the show serverfarm output when the server failed. Please help me.
If you add "failaction purge" to the serverfarm definition the ACE will terminate any sessions bound to a failed rserver. The default is to do nothing - "If you do not configure this command, the ACE takes the real server out of rotation for new connections and allows existing connections to complete. The ACE does not send the connections to a backup server in the server farm or to a backup server farm if all servers in the primary server farm fail." (ACE Command Reference).
thanks for your advice. But the customer doesn't use a backup rserver or a backup serverfarm. He wants to use remaining rservers in the serverfarm:
... - so when the server s1 will fail, the connection for the same user (with cookie for rserver s1) would be load-balanced between remaining servers (s2 and s3). I think that ACE makes it, because I see failed connection for these rservers in the show serverfarm TEST, but the customer doesn't have any output in the browser.
You don't need a backup server or serverfarm to use the "failaction purge" command. Users on the failed rserver will have their session terminated and they'll have to restart their connect. This will then bind to one of the remaining active servers. There isn't an easy way to transparently transfer the session. The point of stickiness is that it is used in a situation where the client and the server maintain a state (e.g. the user's authentication), so unless this state information is replicated between the servers "behind the scenes" it doesn't make a lot of sense to fail the session to another rserver.
I tried to use commands failaction purge and failaction reassign, but the behaviour is same:-( See:
sho serverfarm TEST serverfarm : TEST, type: HOST total rservers : 6 --------------------------------- ----------connections----------- real weight state current total failures ---+---------------------+------+------------+----------+----------+--------- rserver: s1 10.254.254.18:0 8 OPERATIONAL 0 1878 68 rserver: s2 10.254.254.24:0 8 OUTOFSERVICE 0 409 1
The rserver s2 - it is the server where the client sent first request, for now it is out of service and the failures are increasing for rserver s1.:-(
sho serverfarm TEST detail serverfarm : TEST, type: HOST total rservers : 6 active rservers: 1 description : - state : ACTIVE predictor : LEASTCONNS slowstart : 0 secs failaction : PURGE back-inservice : 0 partial-threshold : 0 num times failover : 0 num times back inservice : 0 total conn-dropcount : 0 Probe(s) : HTTP, type = HTTP HTTP_443, type = HTTP
Introduction This article will help you understand the steps on how to
download the UCS licenses from the Cisco Systems website and then
installing it on the UCS. The redacted (blue lines) just covers up
certain numbers for privacy please do not take them...
Introduction This article will help you understand and educate the
customer on how to clear their "expired licenses"
(license-graceperiod-expired) from their UCS-M. If a customer just
purchased a license and needs a step by step guide on how to download
Introduction Prepositioning is a powerful tools on the WAAS platform but
it is not always easy to figure out why your jobs are failing when
trying to retrieve the files.Here is a method that should help you to
figure out the reason why they are not succes...