Pictrivia 9 - What load balancing technique is this?
What's Pictrivia?
It's a trivia presented as a picture - more picture, less words. If you love solving puzzles, racking your brains on brain teasers,you would love this one.
Who can participate?
It's open to all. You can participate on Facebook, Twitter or here. If you are subscribed to our Facebook or Twitter channels, you'll see the message posted on early Saturday mornings pacific time.
How often will you post a trivia?
Once a week on Fridays.
How do I answer?
You can answer by replying to this discussion. If you saw the post on our Facebook page or Twitter, simply comment on the post or reply to our tweet.
What do I get?
Intellectual stimulation, thrill and pleasure of solving a puzzle!
What's the skill level required?
Beginners
When will I know the right answer?
On the following Wednesday.
It's Standard Firewall Load Balancing.
Regards.
Similar Messages
-
now that our iPad is upgraded to ios 7.+, I cannot save images from google to photo gallery the old conventional way. Whats the new technique?
actually with new ios 7
select image from search,
once image has been selected and enlarged you may tap/hold select
then options will appear...then 'save image' -
What is load balancing "stickyness" problem
We are getting this error " Failed to send alert messages to browser " on the interaction centre after saving the ticket followed by the END button.
There is a similar thread with the same issue [CRM 2007-IC Web Interface Message - Failed to send Alert Message to Browser;
in the observations one of the solution for the issue was load balancing "stickyness" problem
If any one can share some knowledge on load balancing "stickyness" problem it is really appreciated and points will be awarded.
regards
Kumar
Edited by: S Kumar on Jun 16, 2011 11:28 AMClustering : is the use of multiple computers to provide a single service.
Load Balancing: Technique implemented to spread "load" between alike computers for service availability via unshared system resources increasing system availability and performance.
Please refer to this link to gather more about :
http://en.wikipedia.org/wiki/Load_balancing_%28computing%29 -
Hi
Using Dell 2U servers running FreeBSD 6, we are very exciting to get some new Xserve for our web needs.
We plan to buy 2 XServe for sharing performances for a huge website and a Xserve RAID.
As MySQL can be master-master replicated on its own, we only want to balance network load coming from the internet. What do you suggest to buy in front of the 2 Xserve ? How to sync files between the 2 Xserve but with manual rsyncs ?
Thank you for your tips
PowerBook 12" + MacBook rev1 Mac OS X (10.4.8) Airport Express / 23" Cinema Display / FreeboxWhat kind of traffic levels are you planning for?
There are various load balancing techniques around ranging from the free to the very expensive, and the inefficient to the highly effective.
At the lowest end of the scale is simple round-robin DNS. You configure your site's address with two IP addresses and the DNS server alternates between the answers. This gives you a crude load balancing option - there's no direct control over which server gets the traffic, levels may be uneven and, worst of all, there's no redundancy in case one server is down - the DNS server will continue to hand out it's IP address. Its advantage, though, is that it's free.
Moving up the scale a little there are various Linux based solutions that can do simple load balancing through its IPTables (or ipchains in older distributions).
I've never used them, so I don't know how effective they are.
At the top end of the scale are load balancing appliances such as those from F5, Cisco, NetScaler and others.
These move up the price chain a fair way but offer far more features, server health monitoring (to make sure the server is able to service the request), advanced load balancing rules to decide which server should handle the request, and multi-gigabit per second throughput.
If you just have a couple of servers, the appliance path may be overkill, although if you expect to grow then it may be something worth considering.
As for the replication question, there are many ways of doing that. At its simplest level, rsync can replicate a directory or filesystem using an efficient protocol that just transfers the differences. It's included in Mac OS X and the man page gives examples of its use. -
Dear Gurus
my database 10.2.0.4 and rac is configured.
database is working fine.actually rac is installed from last 3 years and database is running fine.
But when we see cpu usage on both nodes by command sar -r ,cpu usage on node 2 is less than node1
so my boss is saying that load balancing is not happening.
Could anybody tell me how could i check whether load balancing is fine on both nodes or is it like some master node concept.
Please give some ideas to test the same as in my opinion load balancin is working fine but i don't know how do i check itHi
So, you are trying to achieve uniform CPU usage among the RAC nodes using RAC load balancing techniques?
I am afraid, answer to your problem is not simple. Do you use Services to load balance? If yes, how is the CLB_GOAL and GOAL (run time) is setup for the services? You could potentially use services and load balancing goal to keep the CPU usage uniform.
But, this becomes more complex in connection pool environments. In connection pool, connections are created at an accelerated pace. PMON updates the listener with Load Balancing Advisory data every 3 seconds or so, and if the connections arriving within that 3 seconds, then the listener does not get correct LBA and So, the listener might redirect the connection to the node already heavily loaded. You could control the this damage by introducing delays between the connection pool logins in such a way that PMON has enough time to update the listener.
Even then, In theory, you could have imbalanced CPU usage. For example, let's say that application created 50 connections to instance 1 and 50 connections to instance 2. Now, some part of the application is going to have more CPU usage, then the others. If accidentally, all those connections borrowed by those CPU intensive components, are from one instance, then CPU usage will be higher. I guess, what I am saying is that connection pools do not know underlying instance load and might not distribute the workload.
I will ask the question differently. Were you able to identify exactly what is causing the increased CPU usage in the heavily loaded node? Are you sure that it is the application? If yes, you could potentially configure more services so that you can differentiate the CPU usage at the service level.
Now, the ultimate question is, do you see this imbalance causing performance issue?
Hope this helps,
Cheers
Cheers
Riyaj Shamsudeen
Principal DBA,
Ora!nternals - http://www.orainternals.com - Specialists in Performance, RAC and EBS
Blog: http://orainternals.wordpress.com
Oracle ACE Director and OakTable member http://www.oaktable.com
Co-author of the books: Expert Oracle Practices, Pro Oracle SQL, Expert PL/SQL Practice
Join me: Advanced RAC seminar - see blog. -
Trouble in a SAP system with load balancing
Hello masters:
I had a trouble with workitems that came from a SAP system with load balancing. They doesn't receive the workitem ID in the UWL.
I created another system in the portal as a "dedicated application server", and made the necessary configuration in order to replace the first one, and then my workitems worked fine.
However, I would like to know what appened with the first system created as "system with load balancing". Why this system didn´t pass the workitem ID to the application (web dynpro abap) that execute the workitems?
Thank you.I hope there was some missing configuration in the first case which was not replicating the workitem from SAP Inbox to UWL.
Thanks
Arghadip -
/* Style Definitions */
table.MsoNormalTable
{mso-style-name:"Table Normal";
mso-tstyle-rowband-size:0;
mso-tstyle-colband-size:0;
mso-style-noshow:yes;
mso-style-priority:99;
mso-style-qformat:yes;
mso-style-parent:"";
mso-padding-alt:0in 5.4pt 0in 5.4pt;
mso-para-margin-top:0in;
mso-para-margin-right:0in;
mso-para-margin-bottom:10.0pt;
mso-para-margin-left:0in;
line-height:115%;
mso-pagination:widow-orphan;
font-size:11.0pt;
font-family:"Calibri","sans-serif";
mso-ascii-font-family:Calibri;
mso-ascii-theme-font:minor-latin;
mso-fareast-font-family:"Times New Roman";
mso-fareast-theme-font:minor-fareast;
mso-hansi-font-family:Calibri;
mso-hansi-theme-font:minor-latin;}
This topic has been beat to death, but I did not see a real answer. Here is configuration:
1) 2 x ASA 5520, running 8.2
2) Both ASA are in same outside and inside interface broadcast domains – common Ethernet on interfaces
3) Both ASA are running single context but are active/standby failovers of each other. There are no more ASA’s in the equation. Just these 2. NOTE: this is not a Active/Active failover configuration. This is simply a 1-context active/standby configuration.
4) I want to share VPN load among two devices and retain active/standby failover functionality. Can I use VPN load balancing feature?
This sounds trivial, but I cannot find a clear answer (without testing this); and many people are confusing the issue. Here are some examples of confusion. These do not apply to my scenario.
Active/Active failover is understood to mean only two ASA running multi-contexts. Context 1 is active on ASA1 Context 2 is active on ASA2. They are sharing failover information. Active/Active does not mean two independently configured ASA devices, which do not share failover communication, but do VPN load balancing. It is clear that this latter scenario will work and that both ASA are active, but they are not in the Active/Active configuration definition. Some people are calling VPN load balancing on two unique ASA’s “active/active”, but it is not
The other confusing thing I have seen is that VPN config guide for VPN load balancing mentions configuring separate IP address pools on the VPN devices, so that clients on ASA1 do not have IP address overlap with clients on ASA2. When you configure ip address pool on active ASA1, this gets replicated to standby ASA2. In other words, you cannot have two unique IP address pools on a ASA Active/Standby cluster. I guess I could draw addresses from external DHCP server, and then do some kind of routing. Perhaps this will work?
In any case, any experts out there that can answer question? TIA!Wow, some good info posted here (both questions and some answers). I'm in a similar situation with a couple of vpn load-balanced pairs... my goal was to get active-standby failover up and running in each pair- then I ran into this thread and saw the first post about the unique IP addr pools (and obviously we can't have unique pools in an active-standby failover rig where the complete config is replicated). So it would seem that these two features are indeed mutually exclusive. Real nice initial post to call this out.
Now I'm wondering if the ASA could actually handle a single addr pool in an active-standby fo rig- *if* the code supported the exchange of addr pool status between the fo members (so they each would know what addrs have been farmed out from this single pool)? Can I get some feedback from folks on this? If this is viable, then I suppose we could submit a feature request to Cisco... not that this would necessarily be supported anytime soon, but it might be worth a try. And I'm also assuming we might need a vip on the inside int as well (not just on the outside), to properly flip the traffic on both sides if the failover occurs (note we're not currently doing this).
Finally, if a member fails in a std load-balanced vpn pair (w/o fo disabled), the remaining member must take over traffic hitting the vip addr (full time)... can someone tell me how this works? And when this pair is working normally (with both members up), do the two systems coordinate who owns the vip at any time to load-balance the traffic? Is this basically how their load-balancing scheme works?
Anyway, pretty cool thread... would really appreciate it if folks could give some feedback on some of the above.
Thanks much,
Mike -
For a true load balancing and high-availability OHS, OPMN, and mod_oc4j
i have read this link of Enabling Clustering on oc4j9.0.4 standalone app server
http://www.oracle.com/technology/docs/tech/java/oc4j/htdocs/getstart.htm#1015479
To test the clustering, start up the load balancer by executing "java -jar loadbalancer.jar".
C:\OC4J_EXTENDED\j2ee\home>java -jar loadbalancer.jar
In a future release of Oracle Application Server, loadbalancer.jar will be
desupported. Because of this, we strongly suggest that you discontinue your use
of loadbalancer.jar in this release. Under high loads, loadbalancer.jar may not
function properly. For a true load balancing and high-availability solution,
please move to use OHS, OPMN, and mod_OC4J. For more information, please see
http://otn.oracle.com/products/ias/ohs/content.html
Balancer initialized...
what load balancer should i use for web clustering
<frontend host="balancer-host" port="balancer-port" />
balancer-host=localhost
balancer-port=80
for all nodes i mentioned same host and port in http-web-site.xml.Is it correct?
i completed all the steps and run http://localhost:6666/session/SessionServlet
i hit 3 times
in the different browser http://localhost:7777/session/SessionServlet
instead of coming 4 it starting from 1 only.can i use this loadbalancer.jar or not?
how to mod_oc4j in standalone app server -
Hi,
I have one query regarding balancing load with RAC.
We have some batch processes executed from shell script in Oracle RAC environment where a separate node is used for failover only. There are some issues related with performance. We are looking for a solution to utilize 'idle' node used for failover to resolve this issue.
I want to know whether the below mentioned scenario is possible to achieve with 10gRel2 RAC.
A batch process is running and it is utilizing all resources (CPU) of instance 1, then it is possible by any means that this process be split and distributed to idle instance 2 on another node?
Looking for expert advice.
Regards
AmitHi Amit,
I;ve done lots of RAC implementations, and it;s critical to set-up the primary node accordig to the "type" of table access. For example, cistomer queries on node 1, order queries on node 2, etc. Minimizing pinging is very important!
Staring in Oracle 10g release 2, Oracle JDBC and ODP.NET provide connection pool load balancing facilities through integration with the new “load balancing advisory” tool. This replaces the more-cumbersome listener-based load balancing technique.
I have more notes here:
http://www.dba-oracle.com/oracle10g_tuning/t_grid_tuning.htm
Hope this helps. . .
Donald K. Burleson
Oracle Press author -
Using StoneBeat WebCluster load balancing with WebLogic
Hi,
I have done some testing of WebLogic Server with my company's StoneBeat WebCluster
distributed load balancing software. This might be one more option to consider
as a load balancing solution for WLS. It is advanced in the sense that load balancing
is really dynamic, there are no single-points of failure (distributed architecture)
and there is a very good, configurable test subsystem that runs on each cluster
node to check for overload situations, HW/OS failures, ...
In the initial testing, the WebCluster load balancing works with WebLogic replication,
although there are some cases that need mroe consideration (please see below).
I had to get a patch to WLS6SP1 on NT to make WLS' multicast work when there are
several NICs on the cluster nodes.
However, there is one case which causes problems:
- I have 3 cluster nodes
- P: 2, S: 3 (SessionServlet = 1)
- 2: offline - P: 3, S: 1 (SessionServlet = 2; WebCluster randomly selected a
new node to handle the connection)
- 2: online - P: 2, S: ? (SessionServlet = 3, WebCluster redistributes the load
when a node goes online)
- 2: offline
- P: 3, S: 1 (__SessionServlet = 1__) NB!
The log messages show that when node 2 comes back online it retrieves the replica
from the secondary (node 1) and not from the primary (node 3). After a while (5-6
minutes), node 3 tries to update the replica on node 1. Node 1 considers this
a stale update request and removes the Primary 16... (node name) and then the
secondary for 16... (the replicated object). Then there's a message (still on
node 1) that it is unable to find object 16... Back on node 3 the primary for
16... is removed.
From the WLS6 documentation (under the discussion of using replication with external
HW load balancing solutions) I thought that this case would have been handled:
- it is stated that after the failure of a node, if the HWLB box sends the next
request to a node where there is no replica, WLS is able to retrieve the replica
- to be fair, this is what happens: when node 2 came back online, it retrieved
the replica from node 1 (the secondary) - I suppose that there is an assumption
that if a request arrives to a node without a replica, the primary __must have
failed__
Is there any way to get around this problem?
Admittedly, WebCluster has a problem in that the stickyness of connections is
not perfect: - when a node goes online, a connection that was correctly persisted
(based on either source-ip or source-network address) may be moved to a new node
since the load is redistributed. Our load balancing is very dynamic, but doesn't
maintain a list of who is connected to which node when resistribution takes place
Regards,
Frank Olsen
Stonesoft
Rick,
You may want to look at the Alteon and F5 configuration we have on edocs.
Take a look at the following URLs for a possible solution
http://edocs.bea.com/wls/docs61/cluster/alteon.html#591902
http://edocs.bea.com/wls/docs61/cluster/bigip.html#591902
Chuck Nelson
DRE
BEA Technical Support -
3rd party distributed SW load balancing with In-Memory Replication
Hi,
Could someone please comment on the feasibility of the following setup?
I've started testing replication with a software load balancing product. This
product lets all nodes receive all packets and uses a kernel-level filter
to let only one node at the time receive it. Since there's minimum 1 heartbeat
between the nodes, there are several NICs in each node.
At the moment it seems like it doesn't work: - I use the SessionServlet - with
a 2-node cluster I first have the 2 nodes up and I access it with a single client:
.the LB is configured to be sticky wrt. source IP address, so the same node gets
all the traffic - when I stop the node receiving the traffic the other node takes
over (I changed the colours of SessionServlet) . however, the counter restarts
at zero
From what I read of the in-memory replication documentation I thought that it
might work also with a distributed software load balancing cluster. Any comments
on the feasability of this?
Is there a way to debug replication (in WLS6SP1)? I don't see any replication
messages in the logs, so I'm not even sure that it works at all. - I do get a
message about "Clustering Services startting" when I start the examples server
on each node - is there anything tto look for in the console to make sure that
things are working? - the evaluation license for WLS6SP1 on NT seems to support
In-Memory Replication and Cluster. However, I've also seen a Cluster-II somewhere:
is that needed?
Thanks for your attention!
Regards, Frank Olsen
We are considering Resonate as one of the software load balancer. We haven't certified
them yet. I have no idea how long its going to take.
As a base rule if the SWLB can do the load balancing and maintain stickyness that is fine
with us as long as it doesn't modify the cookie or the URL if URL rewriting is enabled.
Having said that if you run into problems we won't be able to support you since it is not
certified.
-- Prasad
Frank Olsen wrote:
> Prasad Peddada <[email protected]> wrote:
> >Frank Olsen wrote:
> >
> >> Hi,
> >>
> > We don't support any 3rd party software load balancers.
>
> Does that mean that there are technical reasones why it won't work, or just that
> you haven't tested it?
>
> > As >I said before I am thinking your configuration is >incorrect if n-memory
> replication is not working. I would >strongly suggest you look at webapp deployment
> descriptor and >then the config.xml file.
>
> OK.
>
> >Also doing sticky based on source ip address is not good. You >should do it based
> on passive cookie persistence or active >cookie persistence (with cookie insert,
> a new one).
> >
>
> I agree that various source-based sticky options (IP, port; network) are not the
> best solution. In our current implementation we can't do this because the SW load
> balancer is based on filtering IP packets on the driver level.
>
> Currently I'm more interested in understanding whether it can our SW load balancer
> can work with your replication at all?
>
> What makes me think that it could work is that in WLS6.0 a session failed over
> to any cluster node can recover the replicated session.
>
> Can there be a problem with the cookies?
> - are the P/S for replication put in the cookie by the node itself or by the proxy/HW
> load balancer?
>
> >
> >The options are -Dweblogic.debug.DebugReplication=true and
> >-Dweblogic.debug.DebugReplicationDetails=true
> >
>
> Great, thanks!
>
> Regards,
> Frank Olsen
-
Load balancing across 4 web servers in same datacentre - advice please
Hi All
Im looking for some advice please
The apps team have asked me about load balancing across some servers but im not that well up on it for applications
Basically we have 4 apache web servers with about 2000 clients connecting to them, they would like to load balance connections to all these servers, they all need the same DNS name etc.
what load balancing methods would I need for this, I believe they run on Linux
Would I need some sort of device, or can the servers run some software that can do this, how would it work? and how would load balancing be achieved here?
cheersCarl,
What you have mentioned sounds very straightforward then everything should go well.
The ACE is a load balancer which takes a load balancing decisions based on different matching methods like matching virtual address, url, source address, etc then once the load balance decision has been taken then the ACE will load balance the traffic based on the load balance method which you have configured (if you do not configure anything then it will use the default which is "round robin"), then it will send the traffic to the servers which it has available and finally the client should get the content.
If you want to get some details about the load balancing methods here you have them:
http://www.cisco.com/en/US/docs/app_ntwk_services/data_center_app_services/ace_appliances/vA3_1_0/configuration/slb/guide/overview.html#wp1000976
For ACE deployments the most common designs are the following.
Bridge Mode
One Arm Mode
Routed Mode
Here you have a link for Bridge Mode and a sample for that:
http://docwiki.cisco.com/wiki/Basic_Load_Balancing_Using_Bridged_Mode_on_the_Cisco_Application_Control_Engine_Configuration_Example
Here you have a link for One Arm Mode and a sample for that:
http://docwiki.cisco.com/wiki/Basic_Load_Balancing_Using_One_Arm_Mode_with_Source_NAT_on_the_Cisco_Application_Control_Engine_Configuration_Example
Here you have a link for Routed Mode and a sample for that:
http://docwiki.cisco.com/wiki/Basic_Load_Balancing_Using_Routed_Mode_on_the_Cisco_Application_Control_Engine_Configuration_Example
Then as you could see in all those links you may end up having a configuration like this:
interface vlan 40
description "Default gateway of real servers"
ip address 192.168.1.1 255.255.255.0
service-policy input remote-access
no shutdown
ip route 0.0.0.0 0.0.0.0 172.16.1.1
class-map match-all slb-vip
2 match virtual-address 172.16.1.100 any
policy-map multi-match client-vips
class slb-vip
loadbalance vip inservice
loadbalance policy slb
policy-map type loadbalance http first-match slb
class class-default
serverfarm web
serverfarm host web
rserver lnx1
inservice
rserver lnx2
inservice
rserver lnx3
inservice
rserver host lnx1
ip address 192.168.1.11
inservice
rserver host lnx2
ip address 192.168.1.12
inservice
rserver host lnx3
ip address 192.168.1.13
inservice
Please mark it if it answered you question then other users can use it as reference in the future.
Hope this helps!
Jorge -
Linksys RV082 Load Balancing and VPN
We have 3 offices connected by VPN and are currently using RV082 routers for the endpoints. The 2 branch offices have service from a small wireless ISP in the area. We have recently brought DSL in to the main office in addition to the wireless ISP. There are about 20 workstations in the main office, and 2-5 in each of the other offices.
Right now we have the RV082 at the main office configured to use the DSL as the primary service and fail over to the wireless ISP in event of issues with the DSL. What we would like to do now is have load balancing for the main office and have the VPN endpoints run specifically on the wireless ISP (much lower latency than going through the DSL), failing over to the DSL when required. The ideal situation would allow us to specify traffic (ie - web, and email) for a 'preferred' WAN port so that it uses, for example the DSL circuit for email traffic until the DSL circuit goes down, then email traffic is automatically moved over.
I know that you can set rules in the load balancing configuration of the RV082 to direct traffic out a specific port. When I have configured these in the past for testing, the rules were not removed when the connection failed over, so any traffic that had a rule configured for the failed port stopped working.
All of the routers are running firmware version 2.0.0.19-tm.
I'm sorry if this is a scenario that has been addressed before, but I haven't been able to find any answers. If I missed it, please feel free to point me in the right direction. Is it possible to configure the load balancing rules so that they fail over correctly on the RV082? If so, please post steps for me to follow if you can. If this is not possible on the RV082, please make suggestions for which router we should install at the main office to make this scenario work properly.
Thanks for any help!Thank you for the quick response. Perhaps I didn't explain the situation very well in my first post. We are currently using the Smartlink Backup functionality, and it is working correctly.
We would like to move to the Load Balance mode, because this will allow us to take better advantage of having two providers. The Auto Load Balance works fine except for certain sites, like a banking site. I understand that the way around this is to create a Protocol Binding rule, but is there a way to have Protocol Binding rules termporarily removed or modified in the event of a WAN failure, so that traffic will not get stuck being sent out to a service provider that is down? I have tested the Load Balance feature and created some Protocol Binding rules as a test. For example, I bound all HTTP (port 80) traffic to WAN1, then unplugged WAN1 to simulate an outage. All HTTP traffic stopped working after I unplugged WAN1, but all other traffic seamlessly continued to function as if nothing had changed.
Thanks! -
Load Balancing, Tomcat, and SharePoint
I'm a new BusinessObjects customer and am working on getting all of the hardware in place for a new install. The initial plan is to have two BOE servers, and two tomcat virtual servers, with a hardware based load balancer (F5 BIG-IP Switch: Local Traffic Manager 1600 4GB, possibly) in front of the tomcat servers to handle the load balancing.
But, I'm starting to think that it would be a good idea for us to integrate directly in to our SharePoint portal using the SAP SharePoint Integration option.
So, here's my question. If we do that, then our users will get to their BusinessObjects information via SharePoint. So, SharePoint and IIS will be the web server. Will we still have a need for the Tomcat servers? Will SharePoint point to the load balancer, then to Tomcat, and finally to the BOE server, or will it go right to the BOE server, thus negating the need for the tomcat servers altogether, and also even negating the need for the load balancer? If so, can the BOE app servers themselves still be load balanced?
Hopefully this all makes sense - like I said, we're a new customer, so I don't fully understand all of what the servers are doing. I've spent several days searching the forums & the web & reading documentation and haven't come up with an answer yet, so I'm reaching out to you all & hoping somebody can clear it up for me. Thanks!!Thank you for your input, Denis, this does help explain things. Also, thanks for pointing me to that Windows Patterns document. I hadn't seen that one - since we're starting out on 4.0, I've only been looking at documents under the 4.0 folders...
It sounds like we may want to utilize this same load balancer to balance the SharePoint traffic as well, while we're at it. Right now our SharePoint server is clustered using Microsoft Clustering, but not load balanced in any way. Or maybe we just post a link from within SharePoint to the regular old BI LaunchPad and call it a day!
So, if I understand correctly, if we utilize SharePoint, any traffic/processing that the Tomcat server would have handled would now be handled by the SharePoint server, which could potentially be significant. The actual processing of the reports, though, will still get handled by the BOE Cluster, which takes care of its own load balancing, so we'll be fine. Really the only thing the hardware load balancer does is allow the presentation layer to be load balanced - the layers beyond that get load balanced automatically via CMS. Is that all somewhat valid? In the patterns doc, there's Application Servers separate from the BOE Servers. We were just going to have two BOE servers and two Web servers. Where do the application servers fall in, and are they load balanced using the hardware piece?
If we were just concerned about High Availability, it seems like maybe we could more easily use Windows Clustering on those Tomcat servers, and avoid the hardware component altogether. Do the Web servers really get hammered that hard that we need them load balanced? It seems like the BOE servers are the ones doing the heavy resource intensive tasks and we get them load balanced without the hardware anyway, so maybe load balancing the web servers is overkill..... I'm sure a lot of this is tough to answer, I'm just trying to get a feel for it. I want us to have our best performance & be somewhat future-proofed, but don't want to buy things that I don't need to! We're probably looking at a small, 10-20 concurrent user setup for starters, and out to say 200 concurrent max once it's fully rolled out...
Thanks again!! -
Question Cluster/Load balancing
Question about iplanet load balancing/Cluster:
Following discussion are based on iAS C++ engine(kcs).
We have four web servers and two iAS servers:
Web1, Web2, Web3, Web4
iAS1, iAS2
All machines run Solaris 8, web server is iWS4.1 SP6,
Application server is iAS6.0 SP2, and both iAS boxes have
same hardware configuration.
1. What's the best load balancing method for this structure?
Per Server Response Time(Web Connector Driven)
Per Component Response Time(Web Connector Driven)
Round Robin(Web Connector Driven)
User Defined Criteria(iAS Driven)
2. What's the criteria for the kxs engine to choose the kcs
engine to sent request if we set Web Connector Driven
load balancing?
3. If we set iAS driven load balancing, what's the criteria
for the web connector used to choose kxs?
4. We got a problem when run load testing for an AppLogic
in this cluster, one iAS CPU average usage got almost
100%, but the other one is just 70%.
We used Per Server Response Time load balancing method.
Thanks.
Hengsee answers inline
hcao wrote:
Question about iplanet load balancing/Cluster:
Following discussion are based on iAS C++ engine(kcs).
We have four web servers and two iAS servers:
Web1, Web2, Web3, Web4
iAS1, iAS2
All machines run Solaris 8, web server is iWS4.1 SP6,
Application server is iAS6.0 SP2, and both iAS boxes have
same hardware configuration.
1. What's the best load balancing method for this structure?
Per Server Response Time(Web Connector Driven)
Per Component Response Time(Web Connector Driven)
Round Robin(Web Connector Driven)
User Defined Criteria(iAS Driven)
it depends on the characteristics and behaviour of your application
>
2. What's the criteria for the kxs engine to choose the kcs
engine to sent request if we set Web Connector Driven
load balancing?
kxs always does round robin to the kjs or kcs engines. The webconnector
selects the kxs to which to send to.
>
3. If we set iAS driven load balancing, what's the criteria
for the web connector used to choose kxs?
as specified by your criteria in the iAS driven section.
The ias instance will send its current list of preferences for ias
intances it got from the criteria to the webconnector. This information
is dynamic and updated constantly.
>
4. We got a problem when run load testing for an AppLogic
in this cluster, one iAS CPU average usage got almost
100%, but the other one is just 70%.
We used Per Server Response Time load balancing method.
again, this can be a valid result depending on the way your applogics
are written. Are they CPU bound, I/O bound or DB bound? Since individual
components execute differently and you specified to use the average of
those results to determine load balancing this can be a valid result
because differences in execution times of your applogics.
>
Thanks.
Hengregards
Han-Dat
Consulting Project Engineer
iPlanet Professional Services - ANZ
iPlanet e-commerce Solutions
- A Sun|Netscape Alliance
Sun Microsystems Australia Pty Ltd
Maybe you are looking for
-
Hi Can any one help was on the net and all of a sudden my macB pro has no sound and none of my music plays in itunes. went into the system pref under sound it shows that my output is empty and will not un mute1 please help stuck in iraq1
-
How do i set it to yahoo when opening a new tab??
i have windows 7 home premium,, when opening a new tab ,, why is it always blank... how can i change it to be automatically in yahoo or goggle??
-
How to create Location using tcode BBPMAININT
Hi, I wanted to create local plants in EBP/SRM. I was using tcode BBPMAININT to create Location but facing problem. Can any one please list steps to follow to create local plant? Is there any other tcode we can use? Regards, SAP Guy.
-
Some my problem related oracle
hi all i have right down some my problem in oracle please reply me please help with query command line. my database is 45 GB then 1. how to check redo log size and redo log buffer size 2. how to check data file and data file space size 3. minimum req
-
I've recorded all tracks in mono, can I pan them to different speakers?
Hi all, As above. Specifically, I've recorded to parts of harmony and I want to pan one right and one left. Not being a techy, I didn't realise that this couldn't be done if recorded in mono, I thought it only matter how the speakers were set up. I'm