Web Dispatcher not doing the load balancing on the portal

Hi Experts
I am having a production issue where the SAP web dispatcher is not doing the load balancing on the portal.
We have ESS/MSS portal with 1 Message server and 2 Application servers. The Web dispatcher is installed on the message server itself. Here is my Web disp profile file
Profile generated by sapwebdisp bootstrap
unique instance number
SAPSYSTEM = 16
add default directory settings
DIR_EXECUTABLE = .
DIR_EXECUTABLE = F:\usr\sap\<SID>\sapwebdisp
DIR_INSTANCE = .
Accessibility of Message Servers
rdisp/mshost = <hostname>.com
ms/http_port = 8111
#Log and Trace
rdisp/TRACE = 2
SAP Web Dispatcher Parameter
wdisp/auto_refresh = 120
wdisp/max_servers = 100
wdisp/shm_attach_mode = 6
configuration as per SAP note 538405
icm/max_conn      = 7000
icm/max_sockets   = 14000
icm/req_queue_len = 6000
icm/min_threads   = 100
icm/max_threads   = 300
mpi/total_size_MB = 500
mpi/max_pipes       = 14000
wdisp/HTTPS/max_pooled_con = 7000
SAP Web Dispatcher Ports
icm/server_port_0 = PROT=HTTP,PORT=8888
SSL
icm/server_port_1 = PROT=ROUTER,PORT=443, TIMEOUT=60
SAP Web Dispatcher Web Administration
icm/HTTP/admin_0 = PREFIX=/sap/wdisp/admin,DOCROOT=./admin,AUTHFILE=icmauth.txt
wdisp/enable_j2ee_groups = TRUE
wdisp/HTTPS/sticky_mask = 255.255.255.255
In my Web dispatcher Admin page, I see all the three application servers, however the requests are going to only 1 App server. We are using ENd to End SSL configuration for the web dispatcher.
We also have a reverse proxy in the landscape and reverse proxy is forwarding all the requests to the Web dispatcher. In Web disp Admin page>Dispatching Module>SSL End to END dispatching, I see only ONE table entry in the dispatching table and it is our Reverse Proxy.
As all the requests are coming from only one source (Reverse proxy), it seems to me that the sap web dispatcher is forwarding those to the same Application server every time.
Can anyone please advise ?
I also tried to configure logon group in NWA, the web dispatcher is detecting the logon group and all the app servers in the logon group. It still not doing the load balancing.
I would greatly appreciate any help.
Thanks
Viny

Vincent, can you please elaborate more ? Is the web dispatcher not able to recognize stateful and stateless application requests ?
I saw that the procedure for configuring SSL Termination on Web dispatcher is long and complicated and looks like SAP web dispatcher needs to have SSL certificate of its own. As we have no ABAP servers and only Java servers, I can not even create the PSEs using STRUST (as described in SAP help -http://help.sap.com/saphelp_nw70ehp1/helpdata/en/48/99c388d7c46bb9e10000000a42189d/frameset.htm
We already have SSL certificates for Java App servers.
I suppose there should be a way for web dispatcher to identify the incoming requests and forward to appropriate application servers.
Any help is much appreciated.
Thanks
Viny

Similar Messages

Web Dispatcher - Reverse Proxy and Load Balancing

I'm finding limited docs on Web Dispatcher with regard to reverse proxy and load balancing. Are you aware of some recent presentations or docs in this area? The info on help.sap.com is not what I'm looking for.
Thanks.

Hi,
best thing is that you look at your scenarios and test the web dispatcher against each of it, like:
- SSL
- Portal only
- Web Dynpro ABAP / Java
- BSP
- Different backend systems like SRM, MDM
- Several backends with 1 Web Dispatcher
After getting a list of use cases that you can test quite easily (installation of Web Dispatcher is done fast and can be done on a local PC), you can contact SAP Support and ask them about the specific problems and questions you encountered. This way, you'll get the official answer, sometimes they will even inform you about "secret" parameters and options.
As of the reverse proxy functionality: there are several version of Web Dispatcher available that differ from the functionality offered. The latest version - 7.2 - is the one that offers the most, i.e. allows you to create rewrite rules like Apache.
SAP Note 908097 - SAP Web Dispatcher: Released releases and applying patches
br,
Tobias

Site not accessible from the Load balanced web front end server - sharepoint 2010

I have a production environment with 2 WFE's(sp-wfe1 & sp-wfe2), 2 APP's and 2 SQL clustered VM's.
2 WFE's are load balanced using hardware load balancer.
An A-Record(PORTAL) is created in DNS for the virtual IP of the load balancer which points to the 2 WFE's.
A web application is created on the WFE's on port 80.
alternative access mapping is configured and the load balanced record "http://PORTAL" is used under the default zone.
Under IIS I have edited the bindings for the sharepoint site at port 80 and added the HOSTNAME as PORTAL.
Result: The site is accessible from outside the server and works fine.
ISSUE: The site is not accessible within the WFE's(sp-wfe1 & sp-wfe2).
When I browse the site from the WFE's server it ask for the credentials and when I enter the credentials and click OK it ask the credentials again and again and in the end displays a blank page.
Kindly help me in this issue because I am clueless and couldn't find anything helpful on the internet.
Regards,
Mudassar
MADDY-DEV Forum answers from Microsoft Forum

Loop back check.
http://www.harbar.net/archive/2009/07/02/disableloopbackcheck-amp-sharepoint-what-every-admin-and-developer-should-know.aspx

Load balancing from the BI portal

Hi everybody,
We have a BW system with two application servers.
When we use BEx, the load balancing leads the user to the application server that is used less.
When we use the portal, everything goes over the central server.
Does anybody know how we can set up load balancing from the BI portal ?
Kind regards,
Marc van Eijndt

The web dispatcher for your portal should do the load balancing... usually the scenario is the other way round - BEX tends to use the server irrespective of its load and Portal will hit the load balancer and hence lead to better performance ....

Load balancing by the proxy plugin

Has anyone encountered this before:
I have a cluster of two WLS 5.1 servers, hosting servlets that serve web
requests. The requests are proxied through a web server ( I have tried
Weblogic, Apache as well as IIS). I also have a tool that simulates
concurrent web requests and fires them to the proxy server.
As per documentation, as the load balancing while proxying requests to
servlets is round robin, I expect that the requests are uniformly
distributed across the two weblogic servers. But what I see is a bit
different. In one case I fired 15 requests and found that 11 went to first
server and 4 went to the other.
Second time when I fired again 2 of them went to the first server and 13 to
the second one. I would expect that around half of the total requests
should be routed to each server everytime so that there is a proper load
balancing done by the proxy. I have not changed any configuration related
to the default load balancing algorithm. So I expect it is round-robin.
Has anyone encountered this before ? This happens to me irrespective of
which proxy server I use (i.e which proxy plugin I use). Is there some
other configuration required and I am missing something or is there some
inherent problem with the load balancing of the proxy plugins. Any info
would be highly appreciated.
Thanks
Mainak

Could you post this in weblogic.developer.interest.plug-in? This group is for
ejb related questions. Thanks.
Bill
Mainak Datta wrote:
Has anyone encountered this before:
I have a cluster of two WLS 5.1 servers, hosting servlets that serve web
requests. The requests are proxied through a web server ( I have tried
Weblogic, Apache as well as IIS). I also have a tool that simulates
concurrent web requests and fires them to the proxy server.
As per documentation, as the load balancing while proxying requests to
servlets is round robin, I expect that the requests are uniformly
distributed across the two weblogic servers. But what I see is a bit
different. In one case I fired 15 requests and found that 11 went to first
server and 4 went to the other.
Second time when I fired again 2 of them went to the first server and 13 to
the second one. I would expect that around half of the total requests
should be routed to each server everytime so that there is a proper load
balancing done by the proxy. I have not changed any configuration related
to the default load balancing algorithm. So I expect it is round-robin.
Has anyone encountered this before ? This happens to me irrespective of
which proxy server I use (i.e which proxy plugin I use). Is there some
other configuration required and I am missing something or is there some
inherent problem with the load balancing of the proxy plugins. Any info
would be highly appreciated.
Thanks
Mainak

Load balancing in the process chains

HI friends
I have 6 subchains in 1 metachain . we are facing a problem where we have to 2 servers . but when we run the process chains they are occuping all the DIA process in only one server . the other server is empty . so is there a way that we can do the load balancing in the process chains . so that the process chains can occupy all the resources in both servers than only one . please do reply . this is very important . thankyou for all your replies .
*Points will be rewarded *.

Hi,
check this out: https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/com.sap.km.cm.docs/oss_notes/sdn_oss_bc_ccm/~form/handler
additionally some information.
You can specify a server on which the chain will run, but this is not really load balancing. Load balancing must be set up by a basis person.
regards
Siggi

FDM keeps give me DCOM Security Error when I configure the Load Balance

Hello, everyone.
I installed FDM on the Windows 2008 Server.
When I configure the Load Balance Server for FDM.
It keeps give me error:
Unable to create Load Balance Manager Object!
Please verify that the user name, password, and domain name are correct.
However, I am pretty sure the user name and password input are correct.
I have tried the solution provides on Oracle Support. And changes the DCOM component's security settings.
However, the issue is still existed.

If after clearning out the security on the DCOM object and attempting to reset the Load Balance Config the same issue occurs, you will need to un-install FDM using the EPM System Un-installer and reboot. Following the reboot you will then choose to install FDM again, configure FDM using the EPM System Configuration utility, and then reset the FDM configuration items.

Call to the load-balanced SO not through router

Hi all
This in fact not correct - if you make a call to the load-balanced SO,
from within that SO it will not go through the router. You can work around
it quite easily
with a simple Forwarding Dispatcher (design pattern).
Whenever you make a call to your load-balanced SO,
it
would go through the router. That is what I
understand. No matter you make the call from inside
or outside the SO.
Anton van Niekerk
[email protected]

Hi all
This in fact not correct - if you make a call to the load-balanced SO,
from within that SO it will not go through the router. You can work around
it quite easily
with a simple Forwarding Dispatcher (design pattern).
Whenever you make a call to your load-balanced SO,
it
would go through the router. That is what I
understand. No matter you make the call from inside
or outside the SO.
Anton van Niekerk
[email protected]

The load balancer is not provisioned error when attempting to deploy a SP 2013 App

Hello:
I am in VS 2013 with SP 2013 colocated on same Win 2008 R2 server. I opened a SP 2013 App and directly hit F5 and got this error:
Error occurred in deployment step 'Install app for SharePoint': The load balancer is not provisioned. Verify the load balancer was provisioned successfully before retrieving endpoint addresses. When creating the project, the first screen asks for a
site to deploy to, and I just took the default, which was http://<servername> for the colocated server.
I googled this and could not find a solution. Would anyone have any direction?
thank you
Harvey
Regards, Harvey

Hi Harvey Flaisher,
Based on the error message, this seems to be a service application not working issue.
Please check the ULS log or Event log to check in which service application the error occurs, like BCS, secure store service, app management service, then re-provision the service application using PowerShell command, or re-create the service application
from UI to check the result.
Thanks,
Qiao Wei
TechNet Community Support

ConnectionFactory - who does the load balancing

          Consider creating a connectionfactory (with server affinity unticked, load balancing
          ticked and using the message delivery policy of round robin) we then go on to
          create a distributed domain targetted at the cluster of two managed server's (managed1
          and managed2)
          If I create a simple java app that put's messages to that distributed destination,
          using the connectionfactory above, who's responsible for doing the load balancing.
          Does the client create the session knowing that the connectionfactory requires
          load balancing and thus takes the responsibility for it, or does the client just
          put a constant stream of JMS messages to the WLS and the connectionfactory class
          takes responsibility for the load balancing
          Who maintain's the delivery state, the client application or WLS (i.e who's job
          is it to look up the last messages queue destination?)


Hi Barry,
          A JMS client's produced messages are first delivered to the WL server
          that hosts the client's JMS connection. The JMS connection
          host remains unchanged for the life of the connection.
          Once produced messages arrive on the connection host,
          they are load balanced to their JMS destination.
          For more information I suggest reading the clustering
          sections of the JMS Performance Guide white-paper. You can find
          the white-paper here:
          http://dev2dev.bea.com/technologies/jms/index.jsp
          Tom Barnes
          Barry Myles wrote:
          > Consider creating a connectionfactory (with server affinity unticked, load balancing
          > ticked and using the message delivery policy of round robin) we then go on to
          > create a distributed domain targetted at the cluster of two managed server's (managed1
          > and managed2)
          >
          > If I create a simple java app that put's messages to that distributed destination,
          > using the connectionfactory above, who's responsible for doing the load balancing.
          >
          > Does the client create the session knowing that the connectionfactory requires
          > load balancing and thus takes the responsibility for it, or does the client just
          > put a constant stream of JMS messages to the WLS and the connectionfactory class
          > takes responsibility for the load balancing
          >
          > Who maintain's the delivery state, the client application or WLS (i.e who's job
          > is it to look up the last messages queue destination?)
          >
          >
          >
          >

How does Tuxedo 10.0 load balance across the BRIDGE?

Is the load balancing round robin or does work done factored in?

Yes and no. In all honesty, see if this white paper helps answer your question: http://www.oracle.com/technetwork/middleware/tuxedo/overview/ld-balc-in-oracle-tux-atmi-apps-1721269.pdf
Regards,
Todd Little
Oracle Tuxedo Chief Architect

Using the network load balancing from the nodes itself

I have installed a 2 node Sun Cluster 3.2, configured a shared ip resource and attached to it a scalable network aware resource working on the two nodes. I have crashed the process on one of the node in such a way that the cluster could not restart it again
In this status I tried to open a connection from another server and the load balancer always sent the traffic to the node that was up which is as expected...
If I try to open a connection from the node on which the process is failed then I get a connection refused meaning that the load balancer is not working in this circumstance.
Is this a bug/ a mis-configuration/ or just an inherent cluster problem.
Is there a solution to this issue?
Regards
Daniel

To answer your first question, no, there isn't anything you can do.
Here is what my colleague suggested while I was away:
Zone-clusters scalable services still require shared-IP zones, which means requests from one app to another would still bounce back due to loopback. Probably wouldn't help here.
They could isolate the services that must talk to other services into their own failover group on exclusive-IP zones. Other services can be setup as originally planned. But maybe there are too many such "dependent services" for this to be useful. Also, each failover service must have its own IP address.
Finally, can these web services be configured so that it tries multiple addresses. In that case, if the shared address foo for service X bounces back (due to X having crashed on the local node), the app itself would retry with address bar for service X? This allows for uniform configuration across all services, namely:
    - try shared address
    - try node 1's own address (either public or clusternode1-priv)
    - try node 2's own address
You can fine tune it so that configurations on node 1 only use node 2's address as backup, and vice versa. I don't know if that is any help.
As for your second question, the answer is that Solaris Container Clusters allow for consolidation and isolation of clusters onto a single set of nodes. Normal containers don't really allow you to consolidate complete clusters in quite the same way. See http://www.sun.com/offers/details/820-7351.html for more.
Tim
---

How to use the Load Balancer Plug-in to serve multiple domains

In SJSAS8.1 SE/EE the asadmin commands that create and maintain a load balancer configuration operate within a domain. When the load balancer configuration is exported an xml file is created that contains all the information for that domain. To make the load balancer plug-in balance the load for multiple domains, the loadbalancer.xml files can be manually merged to conatin the data that is exported from each domain's load balancer configuration.
For example, 2 domains are created, both having a load balancing configuration. After exporting both configurations using the asadmin export-http-lb-config command, the user would then cut and past the cluster information into the single loadbalancer.xml file that resides under the web server's config directory.
An example of the manually merged loadbalancer.xml file follows:
<?xml version="1.0" encoding="UTF-8"?>
<loadbalancer>
<cluster name="domain1">
<instance disable-timeout-in-minutes="30" enabled="true" listeners="http://localhost:1026 https://localhost:38181" name="i1"/>
<instance disable-timeout-in-minutes="30" enabled="true" listeners="http://localhost:1027 https://localhost:38182" name="i2"/>
<web-module context-root="ab" disable-timeout-in-minutes="30" enabled="true"/>
<health-checker interval-in-seconds="5" timeout-in-seconds="60" url="/"/>
</cluster>
<cluster name="domain2">
<instance disable-timeout-in-minutes="30" enabled="true" listeners="http://localhost:1029 https://localhost:38189" name="i3"/>
<instance disable-timeout-in-minutes="30" enabled="true" listeners="http://localhost:1030 https://localhost:38188" name="i4"/>
<web-module context-root="webservice" disable-timeout-in-minutes="30" enabled="true"/>
<health-checker interval-in-seconds="5" timeout-in-seconds="60" url="/"/>
</cluster>
<property name="response-timeout-in-seconds" value="60"/>
<property name="reload-poll-interval-in-seconds" value="5"/>
<property name="https-routing" value="false"/>
<property name="require-monitor-data" value="false"/>
<property name="route-cookie-enabled" value="true"/>
</loadbalancer>
Hope this helps - Mark

Mark, be my savior, I work for SUN as subcontractor at client site. the only one at site ...so I depend on this forum for solutions........
still having trouble failingover to second instance. I have two AccessManagers behind this loadbalancer.
Here is what I saw......
**************LOGS**********************
[20/Jun/2005:14:22:47] failure (15102): for host 128.114.65.13 trying to GET /amconsole/base/AMA
dminFrame, service-passthrough reports: timed out waiting for request body
[20/Jun/2005:14:22:47] warning (15102): reports: lb.runtime: ROUT1014: Non-idempotent request /
amconsole/base/AMAdminFrame cannot be retried.
So I went and updated the loadbalancer.xml (see at the end of the msg). Now I get a different kind of problem...
**************LOGS******************************
[20/Jun/2005:15:25:18] failure (15295): for host 128.114.65.13 trying to GET /amconsole/base/AMA
dminFrame, service-passthrough reports: timed out waiting for request body
[20/Jun/2005:15:25:18] info (15295): reports: lb.runtime: RNTM3003 : Error servicing the request : NoVal
Here is my loadbalancer.xml file...
<loadbalancer>
<cluster name="cluster1">
<instance name="instance1" enabled="true" disable-timeout-in-minutes="1" listeners="http://idm-test-1.ucsc.
edu:80 "/>
<instance name="instance2" enabled="true" disable-timeout-in-minutes="1" listeners="http://idm-test-2.ucsc.
edu:80 "/>
<web-module context-root="amconsole" disable-timeout-in-minutes="1" enabled="true" error-url="sun-http-lber
ror.html" >
<idempotent-url-pattern url-pattern="/*" no-of-retries="3" />
</web-module>
<web-module context-root="amserver" disable-timeout-in-minutes="1" enabled="true" error-url="sun-http-lberr
or.html" >
<idempotent-url-pattern url-pattern="/*" no-of-retries="3" />
</web-module>
<web-module context-root="ampassword" disable-timeout-in-minutes="1" enabled="true" error-url="sun-http-lb
error.html" />
<web-module context-root="amcommon" disable-timeout-in-minutes="1" enabled="true" error-url="sun-http-lberr
or.html" >
<idempotent-url-pattern url-pattern="/*" no-of-retries="3" />
</web-module>
<health-checker url="/" interval-in-seconds="15" timeout-in-seconds="2" />
</cluster>
<property name="reload-poll-interval-in-seconds" value="60"/>
<property name="response-timeout-in-seconds" value="30"/>
<property name="https-routing" value="false"/>
<property name="require-monitor-data" value="true"/>
<property name="active-healthcheck-enabled" value="true"/>
<property name="number-healthcheck-retries" value="3"/>
<property name="route-cookie-enabled" value="true" />
</loadbalancer>
**************************************************************

Web Services Round Robin Service Load Balancer Event Endpoint Failure

I keep seeing these errors in the UlsTraceLogs:
SharePoint Web Services Round Robin Service Load Balancer Event: EndpointFailure Process Name: OWSTIMER Process ID: 3748 AppDomain Name: DefaultDomain AppDomain ID: 1 Service Application Uri: urn:schemas-microsoft-com:sharepoint:service:9b3095eda69947b299d2f873bbfee5ad#authority=urn:uuid:a01381a61b244525ab4fec30cde9dc5f&authority=https://ApplicationServerName:port/Topology/topology.svc
Active Endpoints: 2 Failed Endpoints:1 Affected Endpoint:
http://WFEserverName:port/9b3095eda69947b299d2f873bbfee5ad/ProfileService.svc
what do these errors mean?

ok, thanks, I'll have a look at that.
Going back to my issue... Since I stopped the User Profile Service on the Application server, now I'm getting these non-stop messages in the log:
SharePoint Web Services Round Robin Service Load Balancer Event: EndpointFailure Process Name: w3wp Process ID: 6088 AppDomain Name: /LM/W3SVC/261708640/ROOT-1-130709594108226406 AppDomain ID: 2 Service Application Uri: urn:schemas-microsoft-com:sharepoint:service:9b3095eda69947b299d2f873bbfee5ad#authority=urn:uuid:a01381a61b244525ab4fec30cde9dc5f&authority=https://ApplicationServerName:port/Topology/topology.svc
Active Endpoints: 2 Failed Endpoints:1 Affected Endpoint:
http://ApplicationServerName:port/9b3095eda69947b299d2f873bbfee5ad/ProfileService.svc
SharePoint Web Services Round Robin Service Load Balancer Event: EndpointFailure Process Name: OWSTIMER Process ID: 8304 AppDomain Name: DefaultDomain AppDomain ID: 1 Service Application Uri: urn:schemas-microsoft-com:sharepoint:service:9b3095eda69947b299d2f873bbfee5ad#authority=urn:uuid:a01381a61b244525ab4fec30cde9dc5f&authority=https://ApplicationServerName:port/Topology/topology.svc
Active Endpoints: 2 Failed Endpoints:1 Affected Endpoint:
http://ApplicationServerName:port/9b3095eda69947b299d2f873bbfee5ad/ProfileService.svc
This time, the messages are referring to the same server - the Application Server. In my original question, I should've differentiated the server names when I pasted the message. Originally the message was referring to the Application
Server and Affected Endpoint was referring to a WFE. I'll edit my original post to make it correct.

Load-balancing in the same IP subnet

Can I use load-balancing in the same IP subnet? I have the servers and client in the same IP subnet. I'd like to load-balance client traffic to server traffic. I also need to load balance traffic between servers. Is possible to configure it only in one VLAN?
For example:
CSS:
interface 4/2
circuit VLAN1
ip address 10.0.0.10 255.255.255.0
service s1
ip address 10.0.0.101
active
service s2
ip address 10.0.0.102
active
service s3
ip address 10.0.0.103
active
service s4
ip address 10.0.0.104
active
owner test
content client
vip address 10.0.0.3
add service s1
add service s2
active
content servers
vip address 10.0.0.4
add service s3
add service s4
active
Cat6500:
interface FastEthernet4/1 - clients
no ip address
switchport
switchport mode access
spanning-tree portfast
interface FastEthernet4/2 - servers
no ip address
switchport
switchport mode access
spanning-tree portfast
interface FastEthernet4/3 - CSS
no ip address
switchport
switchport mode access
spanning-tree portfast
interface Vlan1
ip address 10.0.0.1 255.255.255.0
ip policy route-map pokus
access-list 101 permit tcp any eq 80 any
route-map pokus permit 10
match ip address 101
set ip next-hop 10.0.0.10
Thank you
Roman

yes, it's possible - use trunk with two VLANs (slide 9). or you can use 'transparent' mode (slide 11 - your attachment).
answer to your question (I have the problem to understand why there are two links with the same VLAN on the picture):
on the switch are two port interfaces in *switchport* mode (not trunk). now is it clearly?
result:
both methods (bridge mode with two vlans, or transparent bridge mode) use two vlans. it's on you, which type is for you preferred.
my recommendation is - use first method - one link to CSS with trunk configured in bridge mode (one ip subnet, two vlans, default gw for servers isn't css, but parent router)
martin

Web Dispatcher not doing the load balancing on the portal

Similar Messages

Maybe you are looking for