Load Balance a Citrix Farm

Been approached by our server team. They are having issues with DNS load distibution (won't call it balancing as there's no logic to it) with a 10 server Citrix Farm. It seems to me this would be a good candidate for L3 loadbalancing on a CSS 11500 series.
Anyone have any experience or things I need to consider?
Thanks in advance.
Jim

Check that traffic volumes and patterns are the ones that you have dimensioned the circuits for.
For basic css load balancing configuration kindly refer the following url,
http://cisco.com/en/US/products/hw/contnetw/ps792/products_configuration_example09186a008009438d.shtml
http://www.cisco.com/en/US/prod/collateral/contnetw/ps5719/ps792/prod_end-of-life_notice0900aecd804882de.html

Similar Messages

Load Balancing with citrix netscalar

We currently have SQL 2008 servers (some clustered, some not) where we are looking to have no downtime for maintenance operations. I have been looking into load balancing technologies such as citrix netscalar which has sql server integration with
the product and no downtime capability for failover. I am curious as to why they recommend either replication (real-time changes??) or mirroring and not a cluster (they indicate single point of failure but do they mean that citrix cannot intercept the
failover since windows is controlling it vs mirroring?). Also, has anyone used netscalar (or other load balancing technologies) and what have the successes and issues that people have experienced in their environment.

Hi DBAFan,
Since the issue regards Citrix NetScaler, in order to resolve this issue more efficiently, I recommend you contact support from Citrix, you might get a better answer there.
https://www.citrix.com/community.html
The forum is SQL Server, if your business requirement wants little or no downtime along with data protection, personally, SQL Server AlwaysOn Availability Groups should be considered. There is an article about potential data loss(RPO) and potential
recovery time(RTO) for High Availability and Disaster Recovery SQL Server Solutions.
http://blogs.msdn.com/b/sql_pfe_blog/archive/2013/06/15/an-overview-of-ha-dr-solutions-available-for-sql-server.aspx
Regards,
Sofiya Li
Sofiya Li
TechNet Community Support

Load Balancing on a URL with parameters in it.

Hi,
We have two main Server Farms. I have been asked to load balance to each farm based on the url. The problem:
The url looks like this
https://www.domain.com/test/ci/?par1=Default&par2=main&userRole=userrole&mcId=companyname&par4=somethingelse
The bit of the url for the decision making is "mcId", but as I understand it, I cannot use a "?" in the url text string on the CSS. So, how do i do it ?
Many thanks
Wayne

Wayne,
the documentation is actually incorrect.
The '?' does not prevent the advanced-balance url feature to work.
It just changes where the CSS starts looking for the string.
Check this url for a sample config.
http://www.cisco.com/en/US/products/hw/contnetw/ps792/products_configuration_guide_chapter09186a0080409807.html#wp1115519
Regards,
Gilles.

One-Armed Load Balancing

Can CSS 11000 load balance multiple server farms, using different load balancing algorithms on the same ip subnet and having multiple VIPs in the one-armed configuration.
I know this is not an ideal configuration but have to do it for a relocation project.
Thank yoi

yes you can.
No need for a trunk.
But you have to keep in mind that the CSS must see both sides of a connection.
So, obviously the traffic from the client will hit the CSS vip, but for the server response, you have to make sure it goes back to the CSS.
This can be done with source nating or policy routing.
Gilles.

Srv2008 r2, Load balancing causing sessions to Stack and halt logon

Hello all,
We have currently using Appv 4.6 sp3, across 14 terminal servers. These then have access to 2 app servers and also 2 bkr servers.
We are running a srv2008 r2 environment, and running a windows 7 user experience on the terminal servers.
We are running microsoft load balancing via a farm setup. The member of staff that set they system up has recently left and with limited documentation I am struggling a bit to get my head around why the stacking occurs
The problem we have come across is that the system work ok load balancing works a treat, then all of a sudden a user will come along try log on and will take longer than usual to connect there session. This is causing all users that try connecting
to be stacked behind this slower user logging on and so to a point where there could be 20 + people waiting to log on.
This causes us a huge problem as we can have almost 700+ use at a time on the thin client environment.
Is there a setting that can be set to stop this situation happening, ??
Thanks in advance
Lee

Hi Lee,
Does this issue occurs with all users at a time?
Do you have printer redirection enabled?
If yes, then please try below Hotfix and check the result.
Long logon time when you establish an RD session to a Windows Server 2008 R2-based RD Session Host server if Printer Redirection is enabled
http://support.microsoft.com/en-us/kb/2655998
Hope it helps!
Thanks.
Dharmesh Solanki
Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact [email protected]

Load balance business connectors

Hi experts,
We have 4 business connectors 4.7 version running on 5 different servers, Currently we are facing the problem of frequent restart on all of them. I would like to know is there any way we can load balance so that new Idoc processing request will go to the least used server. Also would like to know if increase in number of IDoc in flow will affect the performance of the business connectors.
Regards,
Suresh

Hi,
Please check the below from this article.
https://social.technet.microsoft.com/Forums/exchange/en-US/749e64d6-9924-4150-bac0-ce4174896fb6/relaying-using-load-balancer?forum=exchange2010
For those who still monitor this thread,
The issue here is that the HLB is not configured to Use Source IP. In the Netscaler world, this is configured on a device or load balancer level and is referred to as USIP. (Use Source IP) If Source IP is not configured, the HLB will pass
the traffic to Exchange using the HLB's interface. You can configure this IP as an allowed IP address on your relay connector, but this is a big security risk as previously stated because all traffic that is routed through the HLB has rights to relay.
Concerning the network guy in PeterN22's post, open connections can be a problem and can cause port saturation if you are using an (*) as a wildcard on your RPC load balancing service. I have had much better luck, across all my load balanced deployments
(Citrix, Kemp, Cisco, etc.), when I use static port assignments on all load balanced services. For example:
SMTP - 25
RPC - 59532
OAB - 59533
HTTPS - 443
POP - 110
etc.
Exchange 2013 load balancing is now stateless and doesn't rely on Source IP, but this same issue with SMTP will occur if Source IP is not used. I know this is a recap of some of the other posts, but hopefully the additional information is helpful for
someone who may come across this post.

Relaying using Load Balancer

Hi
Two node Exchnage 2010 DAG. All server roles on the same boxes.
CAS high availability using Cisco load balancer.
We have some applications they are relaying using Exchange 2010 Hub. They are configured using the vip of the load balancer.
The problem is, that they are unable to relay to external recipients, internal is no problem. Using the Hub directly works fine.
What could be the problem?
Regards
Peter

For those who still monitor this thread,
The issue here is that the HLB is not configured to Use Source IP. In the Netscaler world, this is configured on a device or load balancer level and is referred to as USIP. (Use Source IP) If Source IP is not configured, the HLB will pass
the traffic to Exchange using the HLB's interface. You can configure this IP as an allowed IP address on your relay connector, but this is a big security risk as previously stated because all traffic that is routed through the HLB has rights to relay.
Concerning the network guy in PeterN22's post, open connections can be a problem and can cause port saturation if you are using an (*) as a wildcard on your RPC load balancing service. I have had much better luck, across all my load balanced deployments
(Citrix, Kemp, Cisco, etc.), when I use static port assignments on all load balanced services. For example:
SMTP - 25
RPC - 59532
OAB - 59533
HTTPS - 443
POP - 110
etc.
Exchange 2013 load balancing is now stateless and doesn't rely on Source IP, but this same issue with SMTP will occur if Source IP is not used. I know this is a recap of some of the other posts, but hopefully the additional information is helpful for
someone who may come across this post.

CSS Load Balancing Citrix Terminal server, is ti possible ?

Hi we have to balance a Terminal Server Citrix Server Farm with css, did anyone already realize it? Is there any problem to do it ? Someone told me there is nat problem with citrix metaframe terminal server, has anyone information about it ?
Any help will be greatly appreciated. Many thanks
Max

Stickyness means that once a user is directed to a server through the load balancer, that user will remain on the server he was first load balanced to for the duration of their connection. Otherwise, every tcp connection that a user makes is load balanced to whatever servers are configured. There are severalways to configure stickyness. You can do it via:
-source IP
-source IP and destination port
-text string in a cookie or URL
-SSL session ID
take a look at this document that explains it better than I could:
http://www.cisco.com/en/US/products/hw/contnetw/ps792/products_configuration_guide_chapter09186a0080772d96.html

Load balancing RDS Licensing for Citrix

Hi all,
Currently we are in the middle of upgrading our Citrix farm from 4.5/2003 to 6.5/2008 R2. In relation to TS Licensing, we are using per device CALs. The CALs are split between 2 2003 domain controllers that are published to the forest.
Licenses are checked out from both.
In our test environment for 6.5, I set up the same layout. I have 2 2008 R2 servers with just the RD Licensing role installed and I installed 500 per device licenses on each to start with. The servers are not published but I have specified in
group policy for the 6.5 test environment for the terminal servers to point to the 2 servers. The plan is shortly before we do go live, add all new licenses, publish them to the forest (because there are other apps that need the licenses beyond Citrix),
and retire the old license servers.
Now in the test environment, all licenses are being checked out from the server listed first in the group policy and none from the second server. I was thinking that they would split between the 2 on their own but I must be mistaken.
So here are my questions...
Is there a way to force balancing now while they are only referenced in the GPO for the test farm?
Once they are published to the forest, does that set it to load balance?
If no load balancing occurs in either situation, does that mean that licenses will not be checked out on the second server until all licenses have been checked out on the first server?
Is there a better implementation route I should have followed for 2008 R2?
Thanks!

Hi,
How are things going? I just want to check the status of the issue. If you have any update or concern, please feel free to let us know.
Best Regards,
Aiden
If you have any feedback on our support, please click
here
Aiden Cao
TechNet Community Support

Citrix and Load Balancing

We can use a single Citrix Terminal Server over Netlet without a problem.
Has anyone experience using a Terminal Server Farm instead of a single server?
What is the best approach?
- DNS round robin?
- Citrix configuration, i.e. load balancing?
- Netlet configuration?
- other?
Thanks,
Juergen, Thomas

Thomas,
There is an internal white paper from iPlanet Professional Services that
describes a technique for contacting a Citrix ICA Master browser which in
turn supports load balancing across a MetaFrame/Terminal Server Farm.
The gist of it was that the latest ICA clients support a new protocol
"TCPIP+HTTP" which gets around the UDP problem when contacting the browser.
And since the browser will return a different IP depending on which
MetaFrame server it selects, this was solved by use of the "altaddr" setting
that can be done on each MetaFrame server to set an alternate IP address
that will be forwarded through firewalls.
All of these settings are described in the Citrix Administrator's guide, but
the white paper puts it all together along with examples of how the Netlets
should be constructed.
I have not tried this yet myself, and unfortunately, I do not have the white
paper in electronic form right now to send to you.
Hopefully your iPlanet SE, or other tech support will be able to track this
down and send it to you.
Gary Beshara
BP
"Juergen Maihoefner" <[email protected]> wrote in message
news:[email protected]..
We can use a single Citrix Terminal Server over Netlet without a problem.
Has anyone experience using a Terminal Server Farm instead of a singleserver?
>
What is the best approach?
- DNS round robin?
- Citrix configuration, i.e. load balancing?
- Netlet configuration?
- other?
Thanks,
Juergen, Thomas

2 App servers in farm will be auto failover and load balanced?

Dear all,
We are going to setup following SharePoint farm:
2 WFE
2 APP
2 OWA
2 SQL with shared NAS
We will have hardware load balance in front of 2 WFE. From my understanding, the 2 WFE will detect which APP server is available and send the request there.
Hence, if APP1 is down, all traffic come from both WFE will be directed to APP2.
But I am not so sure about the part about load balance. If all the 8 server instance are working properly, the new request come from WFE2 will be sent to APP1 or APP2? Or it depends on which service application is asking?
In addition, we planned to run all search components on APP servers. But the 2 APP servers are both 16GB memory which is lower than suggested specification. I am considering whether I can assign all the search workload on APP2 while all the other on
APP1.
Thanks.
Mark

Here is a good artcicle on how SharePoint load balances service applications:
http://blogs.msdn.com/b/dtaylor/archive/2011/02/23/sharepoint-2010-service-application-load-balancer.aspx
Yes you can create APP02 as search server but then service apps will not be redundant , so it would be better to run all the service apps on both the servers and if you see any memory issues, then scale it later.
As per
http://technet.microsoft.com/en-us/library/651dba4d-8751-4bd8-9492-f2842b2e1177(v=office.15)#HW_Enterprise , 16 GB is recommended for application server so i think you should be okay unless there is very high load .
Amrita Talreja

Adding a 2nd WFE to a Production Farm - Load balanced using F5 Virtual IPs

Hello all,
In much reading I haven't been able to find a more or less straight answer to this question: I have a small-ish SharePoint farm; 1 WFE, 1 App and a SQL cluster. I need to bring a 2nd web front end into the farm. What I am trying to find
out is, may I install the SharePoint bits and join the new web front end to the farm - without putting the web apps' IPs into my F5 pools - without adversely impacting the farm?
What I'm after is time to test the new web front end by redirecting my browser to it via my hosts file. Once I'm confident all is well, I would then configure WFE2 to have traffic forwarded to it through the F5. Is this a reasonable hope?
Thanks in advance for any advice you might provide

Yes, that will work.
Having an extra WFE that isn't included in your load balancer is actually a fairly common practice when you use it as a dedicated crawl target, there' no impact to having it there unused for a while.
Thanks very much for confirming; I appreciate that!

SharePoint 2010 Kerberos on Load balance farm

I have a SharePoint Load balance farm and my site address is https://sharepoint.com(SharePoint alias creates in ADDS which resolves to virtual IP address VIP), do I need to setup spns for https://sharepoint.com or to all the ip adresses of the webservers
used or to VIP?
Thanks,
D

Hi,
you need to set up Kerberos for the URL your users are typing in the browser and for which you have IIS listening. In your case that is
https://sharepoint.com. This address will be registered with IIS on all load-balanced Servers and the application pool should run under the same service account on all servers.
Regards,
Andrei

ACE to load balance Citrix servers

Hello,
Have anyone configured ACE Modules to load balance Citrix Servers (HTTP) ?
Any special considerations needed?
Many thanks,

HI Javier,
There is one complete design guide available on ciso site.
Kindly go through the below mentioned URL for complete config for ACE to load balance CITRIX as follows:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Data_Center/App_Networking/citrixdg_final.html
You will get othe design guides also which can be very useful:
http://www.cisco.com/en/US/netsol/ns751/networking_solutions_design_guidances_list.html
Sachin Garg

Server Farm load balance

Is it possible to load balance servers that reside from the server side and the request comes from the server farm??
Sent from Cisco Technical Support iPad App

Hello,
Yes, you can but you may need a nat-pool to have the response going back to the ACE. Please send me the running-config or showtech of the Context and let me know the VIP with this issue

Load Balance a Citrix Farm

Similar Messages

Maybe you are looking for