Tricking HttpClusterServlet to weight-balance?

To Whom It May Concern:
          With the WebLogicCluster Web Server Plug-In parameter, I know I have to supply a list of available, clustered servers to run HttpClusterServlet.
          If I duplicate the servers in the list (i.e., server1|server2|server1), will the HttpClusterServlet recognize that and do round-robin on all unique servers in the list, or will it just read the list, one server at a time and be tricked into doing weight-based load balancing?
          Sincerely,
          Anthony

I believe overlaying 2 charts is what I would want to do here.
Every plane has a center of gravity which moves based on how much weight is in certain areas. I have been able to plot the line for the envelope (which looks like the image I posted earlier) based on 5 pairs of coordinates on a scatter graph, and connecting them with a line. These numbers will always stay the same for a particular aircraft (but I want to do this for 6 or 7 aircraft) and it only changes if the something physically changes on the airplane (like new instruments). I wanted to be able to change these numbers easily if something does change down the road
Now, based on the user input values for how much weight is at a specific area (front seat, rear baggage area, etc) I can calculate the CG by getting the total weight and the total moment and plotting the intersect of those two points on the Moment envelope chart.
So yes, I suppose I'm looking to overlay the intersect of these two values on the chart I already created that outlines the moment envelope.
If you look at the example I posted earlier, plot a total weight of 2,367 and a moment of 105.2, follow the red lines to the intersect and since that point is within the Moment envelope, the plane is good to fly.
Hope that helps explain.

Similar Messages

Constructing a weight balanced tree

Hi!
I have some difficult in understanding how to construct a weight balanced tree, given the following data:
a(100),b(1),c(100),d(1),e(100),f(1),g(100)where the numbers represent the probabilities (or accesses).
The algorithm says:
Choose the root so that the maximum of the subtrees is minimal, i.e. minimize max(G(lt(x),G(rt(x)). Then process recursively with the subtrees.
So, I start by this:
a: max(0, 303) -> 303
b: max(100, 302) -> 302
c: max(101, 202) -> 202
d: max(201, 201) -> 201 -> d becomes root
e: max(202, 101) -> 202
f: max(302, 100) -> 302
g: max(303, 0) -> 303
Left Subtree:
a(100), b(1), c(100)
a: max(0, 101) -> 101
b: max(100, 100) -> 100 -> b becomes root
c: max(101, 0) -> 101
Right Subtree:
e(100),f(1),g(100)
e: max(0, 101) -> 101
f: max(100, 100) -> 100 -> f becomes root
g: max(101, 0) -> 101
When I then construct the tree, I get the inverse(!) of a weight-balanced graph!
Greets,
Oliver

Oh, okay... got it... the algorithm only constructs NEARLY optimal weight balanced trees...

HttpClusterServlet Proxy Load Balancing with Multiple Clusters

          How do I load balance JSPs and Servlets using the HttpClusterServlet proxy server
          approach when I have more than one cluster?
          It appears from the docs that there is a 1 to 1 relationship between the non-clustered
          managed proxy server that has the HttpClusterServlet and the cluster (1 proxy
          to 1 cluster).


Preferred method would be domain, domainhash, url or urlhash as described in:
http://www.cisco.com/univercd/cc/td/doc/product/webscale/css/bsccfggd/contrule.htm#xtocid2610122.
do not recommend to use round-robin method for load balancing internet proxies. It creates inefficient usage of the storage and bandwidth, as more proxies can have to cache the same objects.

Local Director Weighted Balancing

Hi,
I have LD with sticky cookie-passive and weighted load balancing over 3 servers.
It seems the weights of the 3 server are not respected.
Here are some displays :
LD_Prod# sh sticky
Virtual Machine(s) Sticky
virtualprod:80:0:tcp 240 cookie-passive tolsessionid
LD_Prod# sh weight
Real Machine(s) Weight Timeout
bfpctol3:80:0:tcp 4 never
bfpctol1:80:0:tcp 2 never
bfpctol2:80:0:tcp 2 never
LD_Prod# sh statistics
Real Machine(s) Bytes Packets Connections
bfpctol3:80:0:tcp 530221148 1673183 46326
bfpctol1:80:0:tcp 651629640 2068285 57703
bfpctol2:80:0:tcp 283775703 918391 26235
Note that bfpctol3 has a weight of 4 (doubles the other two server weights), but the number of connections does not shows this situation.
The three servers had been always up since last clear counters command.
Any idea ?
Thanks in advance.

Keep in mind the weighted predictor makes its decision on current, instantaneous incoming traffic. The more traffic you send it the, better the results. Also, I would use the show real command to see the predictor at work. The show real displays the instantaneous connections to each server. The show statistics is the over all number of connections each has been given.
You might also want to consider using the "loaded" predictor. The loaded predictor assigns a weighted portion of connections to a server before moving to the next.
Link to predictor definition:
http://www.cisco.com/univercd/cc/td/doc/product/iaabu/localdir/ldv42/421guide/42ch05.htm#xtocid856564
I hope this helps!
Gonzalo

Load balancing with HttpClusterServlet

 Hi,
 Could someone plz tell me how HttpClusterServlet achieves
 load balancing?
 Suppose there are 3 servers A,B,C (ordered list) in a cluster.
 When the servlet the receives the request for the 1st time will
 direct it to server A and let B be the secondary. Now, when the
 same client makes a 2nd request the servlet reads the cookie and
 directs the request to server A.
 Where is load balancing been achieved here?
 Or is it this way LB ( load balancing ) is done:
 if any other client (other than the one mentioned above) makes a
 request it is directed to server B (since it is the 2nd server
 in the cluster)
 thanx


It is the second option you mention.
 -- Anand
 "pradeep bhat" <[email protected]> wrote in message
 news:3ca0409e$[email protected]..
 >
 > Hi,
 > Could someone plz tell me how HttpClusterServlet achieves
 > load balancing?
 >
 >
 > Suppose there are 3 servers A,B,C (ordered list) in a cluster.
 > When the servlet the receives the request for the 1st time will
 > direct it to server A and let B be the secondary. Now, when the
 > same client makes a 2nd request the servlet reads the cookie and
 > directs the request to server A.
 >
 >
 > Where is load balancing been achieved here?
 >
 >
 > Or is it this way LB ( load balancing ) is done:
 >
 > if any other client (other than the one mentioned above) makes a
 > request it is directed to server B (since it is the 2nd server
 > in the cluster)
 > thanx
 >
 >

Load Balancing and Clustering

How can I connect to my cluster rather than a single
server in the cluster? I have 3 instances of WLS running on the same machine, and I can connect to these three servers separately (the login.jsp page). However, when I shutdown one of them, the page cannot be displayed, which means that I was connected to the server and not the cluster.
To understand it better following is the configuration details which i am having:
1. Admin Server running on 7001 port.
2. Three server instances(Trade1,Trade2,Trade3) are running on TradeCluster having same ip address (127.0.0.1) and different port number 8001,8011,8021 respectively.
3. Trade application can be accesses via (127.0.0.1:8001 or 127.0.0.1:8011 or 127.0.0.1:8021) separately.
If server Trade1 is getting down in that case 127.0.0.1:8001 is showing "page can not be displayed". I tried to use "HttpClusterServlet" for load balancing but it was not working. How i can access application through ome common URL and how can a configure load balancing.
Can anybody help me find out what I might have missed.
TIA
Sunil

Hi,
we need to specify the cluster information in web.xml like
<web-app>
<servlet>
<servlet-name>HttpClusterServlet</servlet-name>
<servlet-class>
weblogic.servlet.proxy.HttpClusterServlet
</servlet-class>
<init-param>
<param-name>WebLogicCluster</param-name>
<param-value>
hostname1:7736|hostname2:7736|hostname:7736
</param-value>
</init-param>
</servlet>
<servlet-mapping>
<servlet-name>HttpClusterServlet</servlet-name>
<url-pattern>/</url-pattern>
</servlet-mapping>
<servlet-mapping>
<servlet-name>HttpClusterServlet</servlet-name>
<url-pattern>*.jsp</url-pattern>
</servlet-mapping>
<servlet-mapping>
<servlet-name>HttpClusterServlet</servlet-name>
<url-pattern>*.htm</url-pattern>
</servlet-mapping>
<servlet-mapping>
<servlet-name>HttpClusterServlet</servlet-name>
<url-pattern>*.html</url-pattern>
</servlet-mapping>
</web-app>
and deploy the application target to cluster
Prasanna Yalam

Session not Sticky with HttpClusterServlet

Hi,
I am using HttpClusterServlet for load balancing. The cluster is configured to Round-Robin algorithm. As per my understanding, the weblogic implements the Round-Robin with Sticky session.
But the problem I am facing is that the session is not sticky and lost if the application is accessed through the cluster. (There is no problem if I point directly to the Managed instance / If only one managed instance of the cluster is started)
Is there anything o do with weblogic configuration to have sticky session?
Please give me some clue, pointers to move ahead. Thanks is advance.
Regards,
Sudhakar P

Hi All,
By default HttpClusterServlet will manage the sticky session by "HttpSession" object. If you are getting any problem with that, we probably need to debug more.
1) Can you give me some setup details how you have set it up ..
2) Also if possible, use Fiddler and get me the fiddler output.
3) Why do you think that it is not doing sticky session,are you getting flipped to another node and session data is not avaible.
Also some traces of logs when it happen...
Thanks

Weight Watchers Mobile on Centro

Hi, Any tips/tricks for getting Weight Watchers Mobile to work on the Centro? If nothing else, at least the points calculator. -Thanks
Post relates to: Centro (Sprint)

Hello and thank you for using the Palm Help Forums!
I am not too sure exactly but you might be able to find a program that will fit your needs at www.freewarepalm.com. I hope this information is useful...
-Pat

Server "Pinning"

We know that the WebLogic HttpClusterServlet handles load
 balancing on initial contact with a client. What about subsequent
 contacts with the client after a session Id is issued and a session
 is established? Does WebLogic pin all subsequent calls from a
 client to the server that originally served him/her? If so, is there
 some way to disable this behavior and have WebLogic load
 balance all calls from a client? Any light shed on this issue is
 appreciated?
 Thanks,
 Ted B.
 San Diego


Yes it will failover.
 Tiffany wrote:
 > Thanks for replying. What happens if the pinned server fails. Does the
 > session fail over to a working server or is it just lost forever?
 >
 > Ted
 >
 > "Prasad Peddada" <[email protected]> wrote in message
 > news:[email protected]..
 > > Ted Barbusinski wrote:
 > >
 > > > We know that the WebLogic HttpClusterServlet handles load
 > > > balancing on initial contact with a client. What about subsequent
 > > > contacts with the client after a session Id is issued and a session
 > > > is established? Does WebLogic pin all subsequent calls from a
 > > > client to the server that originally served him/her? If so, is there
 > > > some way to disable this behavior and have WebLogic load
 > > > balance all calls from a client? Any light shed on this issue is
 > > > appreciated?
 > > >
 > > > Thanks,
 > > >
 > > > Ted B.
 > > > San Diego
 > >
 > > If your session is established, it will try to go to the same server.
 > There
 > > isn't anything currently to disable this behavior sorry.
 > >
 > > --
 > > Cheers
 > >
 > > - Prasad
 > >
 > >

WL Cluster configuration

 Does anyone know of a WL Cluster comprehensive 'step-by-step' manual or something of that nature? BEA web site provides just a little more than definition of terms.
 Thank you.


 I've been struggling to get a simple WLS 6.0
 clustering solution to work for a little
 while now. I can get the various servers to
 register with the administration service OK,
 but I have had no joy in successfully setting
 up an HttpClusterServlet to load balance/fail
 over/do anything at all. Can you also send me
 your collection of docs by e-mail, thank you.
 Any pointers on setting up an HttpClusterServlet
 would also be appreciated.
 John.
 "Cameron Purdy" <[email protected]> wrote:
 >Unfortunately, the docs on the web site are about all that there are. Once
 >you read through them all about 10 times it starts to make sense. I'm
 >sending you a collection of cluster docs (via email) that I've collected.
 >
 >Do you have any particular questions?
 >
 >--
 >Cameron Purdy
 >Tangosol, Inc.
 >http://www.tangosol.com
 >+1.617.623.5782
 >WebLogic Consulting Available
 >
 >
 >"Gennady Shapiro" <[email protected]> wrote in message
 >news:3a704039$[email protected]..
 >>
 >> Does anyone know of a WL Cluster comprehensive 'step-by-step' manual or
 >something of that nature? BEA web site provides just a little more than
 >definition of terms.
 >>
 >> Thank you.
 >
 >

Http ClusterServlet/SSL woes

Heya,
I tried posting this in the clustering newsgroup to no avail, so I thought I'd try here, too. My problem's kind of confusing, so I'll try and break it down. I'm hoping someone recognizes the pattern of my errors!
I am using Weblogic 8.1sp5. I have a setup of 1 admin server, several clusters, and a managed server using the HttpClusterServlet to load balance. I've been tasked to force SSL if the user comes in on the non-SSL port. On our QA environment, I'm using the Demo certs for this experiment.
1. Each of the application EARs have a WAR with a transport-guarantee of CONFIDENTIAL. Great if you hit the members of the cluster directly, but I found that going through the HttpClusterServlet it will stay non-SSL. I added the transport-guarantee to the HttpClusterServlet's web.xml, and it seems to switch to SSL. Problem is, I'm seeing a "backlog" of requests in the Managed Server's log and its access log shows a double hit for each request - one from the originating client IP and one from the Managed Server's IP.
2. I decided to try and shut off the non-SSL port on my clusters' Managed Servers. The cluster will come up, but the Admin Server gets a "Connection Refused" and any attempts to go through the HttpClusterServlet gets me a 403 error! I can still hit the members of the cluster through their IP/port.
Does anyone have any suggestions?? I've been looking/googling/etc and haven't found any hints on how to start solving this..
Thanks,
Anna

Hello Pavel
Thanks for the response but we managed to find the solution to this a while ago. I have not updated this post with the fix yet as it was such a problematic issue I decided to write a blog on it. the full story on the problem will be published on my blog in the next week or so: http://dev2dev.bea.com/blog/hoos/
The fix was to change the server SSL connection details from "Certificate requested but not enforced" to "client certs not requested". This problem only seems to affect WLS8.1sp4 in combination with out certificates which do not specifiy usage.

Hussein Badakhchani

Logic Express Overload Issues on 2007 MBP

So, i've been having a lot of trouble with overload issues on my 2007 MacBook Pro. It's really got me confused because the problem seemed to just start occurring recently on fairly simple projects. My macbook has 2.2 core duo processors, 4gb rams, ~60 gigs of free HD space, and I'm running projects from an external drive. I am aware of the tricks for optimizing projects (balancing processor thread loads and whatnot) so please don't point me to that section of Logic's manual. I just have no idea what the problem is, especially since I used to run fairly large projects in the past with very little free HD space and somehow got away with it.

Repair the Hard Drive and Permissions
Boot from your OS X Installer disc. After the installer loads select your language and click on the Continue button. When the menu bar appears select Disk Utility from the Utilities menu. After DU loads select your hard drive entry (mfgr.'s ID and drive size) from the the left side list. In the DU status area you will see an entry for the S.M.A.R.T. status of the hard drive. If it does not say "Verified" then the hard drive is failing or failed. (SMART status is not reported on external Firewire or USB drives.) If the drive is "Verified" then select your OS X volume from the list on the left (sub-entry below the drive entry,) click on the First Aid tab, then click on the Repair Disk button. If DU reports any errors that have been fixed, then re-run Repair Disk until no errors are reported. If no errors are reported click on the Repair Permissions button. Wait until the operation completes, then quit DU and return to the installer.
If DU reports errors it cannot fix, then you will need Disk Warrior and/or Tech Tool Pro to repair the drive. If you don't have either of them or if neither of them can fix the drive, then you will need to reformat the drive and reinstall OS X.
The error you mention can be fixed by Disk Warrior.

New glass or new camera?

Hi Aperture users,
I have some cash saved up and want to spend it wisely: Top priority is improving the sharpness of my event shots. This question relates to photo kit, but as I use Aperture in processing and you're all photographers, your input would be greatly appreciated.
Current Kit:
D90
Nikon 18-105mm Kit
Nikon 35mm 1.8
Sigma 10-20mm
SB900 flash
I use 'edge sharpen' on virtually every image. So, aside from a rethink of how I'm shooting and the setting's i'm using, I'd like your advice on my next purchase.
At some stage I'm going full frame, so I was thinking about getting the Nikkor 24-70mm 2.8
In your opinion would this be sharper than my 18-105 kit?
Its just about in budget, and I imagine would serve me well for years to come.
Your thoughts greatly appreciated. Apologies for veering slightly away from Aperture, but if you have any considerations from that end that you think would help do let me know.
Thanks,
Rich

Probably not the best place this, as the subject can get highly complexed. Aperture can help you make an informed decision about what to buy. When i decided it was time to spend some money, i used automator to pull out all kinds of metadata to see what various settings were being used most.
Its hard to give specific advice, but here are a few things.
Even a cheap prime will knock an expensive zoom in to a cocked hat, zooms are popular, because they get the shot the prime just can't get, so no image, is one disadvantage of primes, but this depends on the subject matter. Zooms pull lots more air, than a prime, focus is slower and harder on the mechanism.
More relative to Aperture is the image sharpening. I hardly ever find it necessary to sharpen an image. Maybe just my taste.
Plus the reality is you can't sharpen an image once it has been captured. Lots of people bang on about image noise. Any part of an image that does not form as part of the original scene, is noise, so sharpening is adding more noise. At its, best and done well, it can give the illusion of a sharper image, but, great care is needed. If you shoot raw, the in camera applied sharpening can be turned up or down in Aperture, and this is very subtle. Definition on the other hand will only increase the contrast between pixels minimally and this to me is also image friendly. The sharpening tools hidden in the drop down menu are best used with great care, or avoided, as they can be very aggressive.
Camera shake is the big killer of image quality, and in tests i did some time ago with a tripod, i could not see a difference between two images by visual inspection, one shot normal gentle release button pressure and the other shot with the mirror locked up. When i looked at the point source i made for the test movement was just perceptible in the the form of mirror disturbance.
Focus, you often see pros, focus on the face obviously with the centre point, then tilt the camera down to get more subject in frame, instantly moving the focal plane to behind the subjects head.
Conclusion, all that is a load of rubbish because a good photograph beats a technical masterpiece, any day.
Except if the later is also a good photograph. The trick, is getting the balance between the two.
So it's what you do with it as much as what you have got.
Allan

HttpClusterServlet Load Balancing Problem

          Hi,
          I have a problem with LoadBalancing and HttpClusterServlet.
          I have 2 boxes, with a clustered application, deployed on 2 managed servers
          box 1 with adminServer and and a managed server (managed1)
          box 2 with second managed server (managed2)
          on box 1 I have another managed server, which hosts the HttpClusterServlet application
          The cluster is made up of managed1 & managed2, and httpClusterServlet is configured
          to send requests to boths the managed servers in the cluster.
          When I run some tests against the httpClusterServlet web application to see how
          it load-balances requests, I can see that all the requests are redirected just
          to managed1,
          on the same machine.
          If I stop managed1, and run some tests, I can see that httpCluster every time
          tries to contact managed1 first, and since it's down, sends the requests to managed2(handles
          failover correctly).
          If I move the httpClusterServlet application to another box, let's call it box
          3, and re run the tests, it will load-balances the requests between managed1 (box1)
          and managed2(box2) correctly.
          I would like to know if httpClusterServlet has some kind of "server affinity"
          logic (?) or I'm missing something in the cluster configuration (even if I can't
          see clearly how the cluster configuration could affect things, since HttpServlet,
          which is not in the cluster, just should route
          requests to the specified servers in round-robin fashion, I guess)
          Someone in previous posts pointed out http session issues, is it true? and so,
          why?
          Sorry for the far too long post,
          thanks in advance,
          Thomas


Only from load balancing point of view cluster is not really needed.
apart from session replication and failover capabilities, few more benefits are there with clustering only if you are using a web server also.
Stuck threads can be due to lot's of problems like application issues, load issues etc, So first make sure you are getting stuck thread on both servers or only on one, if on one, then in this case your load balancing not working properly and sending more load to that server, and if you are getting stuck threads on both then try to add one more managed server.
Regards
Mukesh Negi
http://weblogicserveradministration.blogspot.com

LACP with Weighted Load Balancing

Hi,
I am trying to figure out how to use weighted load balancing (WLB) with LACP in Nexus 5K. Please can you give any duidance on this because the documentation I found so far is not helpful.
Basically I have a port channel cosisting of two physical 1G ethernet ports and one backup server connecting with two remote SAN hosts over this port channel. Unfortunately the two remote SAN hosts have similar even mac and ip addresses. Thus ordinary source ip/mac load balancing puts them on the same link in the port channel. I want to apply a weight to try to distribute this load.
Many thanks
Sankung
PS: Ultimately, I am getting a 10G NIC for the backup server but in the meantime want to explore this WLB possibility.

advice : get the apache trace dump to find out what stack it is in. I think you must open a TAR .
The error possibly coming from mod_osso ?

Tricking HttpClusterServlet to weight-balance?

Similar Messages

Maybe you are looking for