Least loaded node load - balancing

Similar Questions

• Load balancing between nodes in a cluster analytic provider service

  Hi all

  -First a bit of background on my architecture. My EMP environment consist of 3 servers Solaris:

  Server1: Foundation Services + APS + EAS WLS Server Admin

  Server2: Foundation Services-APS-EA

  Server 3: Essbase Server + server Essbase Studio

  All services are deployed on a single domain. We have a load balancer, sitting in front of Server1 and Server2 that redirects the request according to the availability of services.

  -Consider the APS:

  We have an APS cluster "AnalyticProviderServices" loaded their AnalyticProviderServices1 on Server1 and AnalyticProviderServices2 deployed on Server2.

  So I connect to the APS and connect as User1. Say that the balancer load decides to send my request to server1, so my request are then managed by APS on Server1. Now if APS on server1 is reduced, all requests for APS on server1 are redirected by weblogic to APS on server2.

  Now ideally APS on server2 should say "hey I see what APS on server1 is down so I will take your session, where it was stopped." So I wait for the 2nd APS node in the cluster to tale my session. But this does not happen... I need to log in again when I hit refresh in excel I get the error "Invalid session...". Please log in again". When I opened EAS, according to me, that I was connected with a new session ID. It seems that the cluster nodes are simply load - swing and are not smart enough to take a node that failed, sessions where she had stopped.

  Is my understanding correct or do I have to configure something to do?

  Thank you

  Kent

  Yes, the session will be lost.

  See you soon

  John

  http://John-Goodwin.blogspot.com/

• Configuration of AppsTier (CFP, load balancing) of nodes R12 Multi with RAC 2 nodes

  Dear all,
  
  I installed and configured applications R12 with Oracle RAC 2 nodes. (APP1, DB1 and DB2).
  
  Now, I need to add a layer of application (APP2) to configure high availability, balancing load and PCP.
  
  So, I cloned (adcfgclone.pl appsTier) application for APP2 layer (not shared, the two file systems are different). When I run the application it doesn't knot APP2 App1, more more simultaneous manager works only on a node (can be due to APPLCSF does not point to the mounting NFS share folder), although I've defined managers primary school-children also.
  
  If I run the autoconfiguration of APP1 application service will only work on APP1 and even if I do for APP2, then application works with APP2. I need to have configuration CFP, balancing the load (via DNS if possible) and the high availabllity for the Application tier.
  
  When I start the application, it will start OPMN and APPS Listener on the nodes, but simultaneous manager does not start on both, to a single node.
  
  Is this requires having both APPS name of the node is the same?
  
  Thank you very much
  
  Ravi Purbia
  
  Published by: Purbia delighted on June 9, 2009 18:04
  
  Published by: Purbia delighted on June 9, 2009 18:05
  
  Published by: Purbia delighted on June 9, 2009 18:07

  Ravi,

  If I run the autoconfiguration of APP1 application service will only work on APP1 and even if I do for APP2, then application works with APP2.

  Most likely, it's something with the profile option values, then run utility adchkcfg to get an html report that lists all files and profile options that change when you run AutoConfig on each node.

  I need to have configuration CFP, balancing the load (via DNS if possible) and the high availabllity for the Application tier.

  For load balancing, you must perform the steps according to the guidelines of this document.

  Note: 380489.1 - using Load-balance with Oracle E-Business Suite Release 12
  https://metalink2.Oracle.com/MetaLink/PLSQL/ml2_documents.showDocument?p_database_id=not&P_ID=380489.1

  For the installation of the CFP, please see:

  Note: 388577.1 - using Load-balance with Oracle E-Business Suite Release 12
  https://metalink2.Oracle.com/MetaLink/PLSQL/ml2_documents.showDocument?p_database_id=not&P_ID=388577.1

  When I start the application, it will start OPMN and APPS Listener on the nodes, but simultaneous manager does not start on both, to a single node.

  Any errors in the log file of CM?

  Is this requires having both APPS name of the node is the same?

  It is not necessary.

  Kind regards
  Hussein

• Best way to HTTP in OSB load balancing

  Hello world
  
  We have a cluster of OSB and we must load balance HTTP requests on managed servers. Looking for info on OSB in load balancing, I found that there are essentially two options: use a hardware load balancer or a software solution like Weblogic HttpClusterServlet. For the moment, we have without balancer material available so we will have to take the option of software. I found a few articles on the configuration of HttpClusterServlet as http://redstack.wordpress.com/2010/12/20/using-weblogic-as-a-load-balancer.
  But I have a question for this configuration. If we use a managed server as a proxy HTTP requests between OSB as managed servers, what would happen if the server goes down? I think that one of the main objectives of a cluster deployment is to avoid a single point of failure, but with this configuration, all requests depend on the availability of the managed proxy server.
  
  Could you recommend a configuration of implementation of OSB load balancing?
  
  Thank you in advance,
  
  Daniel.

  Load Balancing in a cluster for http requests can be made using at least 4 different ways:

  (1) - use a hardware load balancer like F5 BigIP LTM
  (2) - use a web server with the plugin from weblogic to the cluster frontend
  (3) - use weblogic with HTTPClusterServlet
  (4) - use the DNS round robin - it works if you managed servers that run on 2 machines (say mach1, mach2) but on the same port. Client HTTP use hostname "mach" to access the URL and the dns has a resolution of names alternating Mach mach 1 and mach2 IP addresses...

  All options except (1) only achieve load balancing and not automatically failover all instances... Balancing load material has the additional feature to probe [sending periodic pings to targets], by which it can detect if the target resource is alive and do not send traffic to other nodes who are alive... That's why the hardware load balancers are worth their investment...
  other options may work if the client is encoded to a retry on failure... so on 2nd or subsequent attempt, the routing is done in the machine that is alive...

  For the options (1), (2)) and (3), you also need a redundancy of the system (material of load balancer, web server or weblogic) to avoid the single point of failure for load balancing... Balancers support material are usually deployed in redundant pairs to achieve...

  Published by: atheek1 on 11/22/2011 15:31

• Hi ALL, did any attempt on the virtual computer NETWORK load balancing using HYPERV on UCS blades

  I try to configure the CASE server cluster by using the Unicast NLB on the virtual machine on different blades on the UCS, it works for awhile, then he abandoned packages.

  I heard that this screenplay of unicast is not supported in the UCS when she used END-host mode in the fabric interconnet...? any attempted before.

  Would it, I use the multicast mode is that something needs to be done on the FBI62020 or the LAN switch upstream. ??

  Header note I found on the implementation of UCS for mulitcast NLBL:

  Microsoft NLB can be deployed in 3 modes:

  Unicast

  Multicast

  IGMP multicast

  For series B UCS deployments, we have seen that the multicast and IGMP multicast work.

  IGMP multicast mode seems to be the more reliable deployment mode.

  To do this, the monitoring settings:

  All NLB Microsoft value "Multicast IGMP" nodes.  Important!  Check ths by logging into EACH node independently.  Do not rely on the MMC of NLB snap.

  An IGMP applicant must be present on the VLAN of NLB.  If PIM is enabled on the VIRTUAL LAN that is your interrogator.  UCS cannot function as applicant IGMP.  If an interrogator of functioning is not present, NLB IGMP mode will not work.

  You must have a static ARP entry on cheating it upstream pointing IP address Unicast NLB on the multicast MAC address NETWORK load balancing.  This need will set up, of course, on the VLAN of the NLB VIP. The key is that the routing for the NLB VLAN interface must use this ARP entry as a unicast IP ARP response may not contain a multicast mac address. (Violation of the RFC 1812)  Hosts on the NLB VLAN must also use the static entry.  You may have several entries ARP.  IOS can use a function of 'alias' of ARP. (Google it.)

  How Microsoft NLB works. -The truncated for brevity Mac addresses.

  TOPOLOGY OF NLB MS

  NETWORK VLAN 10 = subnet 10.1.1.0/24 IP load balancing

  VIP = 10.1.1.10 NETWORK LOAD BALANCING

  Arp entry static switch advanced IP 10.1.1.10 upstream to MAC 01

  NLB VIP (MAC 01, IP 10.1.1.10)

  NODE-A (AA, MAC IP:10.1.1.88)

  NŒUD-B (MAC BB, IP:10.1.1.99)

  Using the IGMP snooping and interrogator VLAN snooping table is filled with the mac NLB address and groups pointing to the appropriate L2 ports.

  MS NLB nodes will send the responses of IGMP queries.

  This snooping table could take 30 to 60 seconds to complete.

  Host on VLAN 200 (10.200.1.35) sends traffic to NETWORK VIP (10.1.1.10) load balancing

  It goes of course to VLAN 10 interface that uses the static ARP entry to resolve to address MAC 01 VIP NETWORK load balancing.

  Since it is a multicast frame destination it will be forward by the IGMP snooping table.

  The framework will arrive at ALL NLB nodes. (NŒUD-A & NŒUD-B)

  NLB nodes will use its load balancing algorithm to determine which node will manage the TCP session.

  Only one NLB node will respond to this host with TCP ACK to start the session.

  NOTES

  This works in a VMware with N1k, standard vSwtich and vDS environment. Where surveillance IGMP is not enabled, the framing for VIP MAC NETWORK load balancing will be flooded.

  NLB can only work with TCP-based services.

  As stated previously mapping an IP unicast to a multicast mac address is a violation implied by RFC 1812.

  TROUBLESHOOTING

  Make sure your interrogator is working. Just to clarify that this does not mean that it is actually at work.

  Wireshark lets check that IGMP queries are received by the NLB nodes.

  Make sure that the ARP response works as expected.  Once Wireshark again is your friend.

  Look at the paintings IGMP snooping. Validate the L2 ports appearing as expected.

  CSCtx27555 [Bug-preview for CSCtx27555] Unknown multicast with destination outside the range MAC 01:xx: are deleted. (6200 FI fixed in 2.0.2m)

  IGMP mode not affected.

  CSCtx27555    Unknown multicast with destination outside the range MAC 01:xx: are deleted.

  http://Tools.Cisco.com/support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCtx27555

  fixed in 2.0(2m)

  Solution: Change the NLB mode of operation of "Multicast" to "multicast IGMP', which modifies balancing load NETWORK VIP MAC at 0100.5exx.xxx Beach, allows to transfer occur as expected.

  Q: and if I switch to switch mode, which means all of the profile and the settings on the servers are completely exhausted and I need to recreate them. ???

  A:Cisco Unified Computing System Ethernet switching Modes

  http://www.Cisco.com/en/us/solutions/collateral/ns340/ns517/ns224/ns944/whitepaper_c11-701962.html

  -There is no impact on the configuration, you have done service profiles.  they will continue to work as expected.  Mode selector has the FI behave more like a conventional switch.  Most notable is that Spanning tree will be activated and if you have several uplinks yew, tree covering weight will begin to block redundant paths.

  You need to review your topology and what impact tree covering weight.  Generally, we at the switch port upstream defined as "edge master", you want to delete this line.

  For pre-production and laboratory environment, PDI can help qualified with the planning, design and implementation partners.  Given to review the IDP site and open a case if you need more detailed assistance.

• To balance the load or not load balance? ISE and F5 Big IP

  My team is currently negotiating to put our two devices of ISE (nodes PSN) behind our F5 load balancing deployment.

  Our network is relatively small (5K users) with a small deployment wireless (Cisco 4 controllers with 300 access points). Growth of the network should remain relatively small in the years to come.

  We will deploy Dot1X wired, followed by reclamation and the assessment of posture. (BYOD isn't an option).

  On the one hand, the characteristics of Big IP could make it easier for us to perform load balancing, maintenance and troubleshooting.

  On the other hand, the Big IP adds another layer of complexity in an already complex deployment. We already have the ability to balance the load of the switches themselves. Load balancing for the wireless should not be a problem that our deployment is very small and I expect to stay that way. Given the size of my environment, it seems to be relatively little to gain for the additional efforts and the potential pitfalls.

  Someone at - it care to share their honest opinion on this issue?

  Thank you

  Phill

  Load balancers are elegant and do their job well when it comes to distribute the load between servers. You already have a then I would use it if you have technical skills to set it up.

  That being said, if your team is not 100% comfortable with F5 then you should certainly jump. Instead, you can configure your WLCs to use as a primary and secondary Radius Server node #2 node #1 , then your switches use Node #2 as primary and secondary node #1 .

  I hope this helps!

  Thank you for evaluating useful messages!

• Client based in 12 G load balancing

  We have a JDBC configuration against an address scan with client load balancing enabled.

  Example:

  JDBC:Oracle:Thin:@(Description=(LOAD_BALANCE=on)(Address=(Protocol=TCP)(Host=xxxxxxxxxxxx.de)(port=xxxx))(CONNECT_DATA=(service_name=XXXXXX)))

  What I understand of Diference between Client-side and Server Load Balancing If you do not use the server load balancing, you can bypass by connecting the service that identifies a particular node of RAC name.

  
  

  My question is, if you have a connection pool object this configuration and if the name service maps node to goes down, what happens to the connection connection pool?

  
  

  Issue 1) Don't scan address switches to the available according to the name of the service node, then the JDBC driver on the client must file all embusked connections and reconnect the available node? or is all the agnostic client failure of node on and all the old connections available in the pool are available for use?

  
  

  
  

  The client side or server balancing side basically works for the same purpose. The difference is that you do not have all the customers to reconfigure then change something in the environment, such as adding or removing a node such as the side server.

  Answer your question...

  If you mean the shared server architecture by connection pool, the shared server session connected to node 1 will die and everyone shares this session should log. Failover is not automatic, you must specify the failover clause in the JDBC or TNS connect string:

  TEST_TAF =

  (DESCRIPTION =

  (ADDRESS = (PROTOCOL = TCP)

  (HOST = rac - scan.example.com) (PORT = 1525))

  (CONNECT_DATA =

  (SERVICE_NAME = test)

  (FAILOVER_MODE = (TYPE = SESSION) (METHOD = Basic))

  ) )

  Or if you use failover and load balancing services you set failover described below:

  for 12 c: srvctl $ add orcl db-test - oel6vm1 favorite-available oel6vm2 - tafpolicy BASE - failovermethod SESSION service service - failoverretry 5 - failoverdelay 60

  for 11g: $ srvctl add service d orcl if test - r oel6vm1 - oel6vm2 BASIC EI SESSION m - z 5 AW 60

  Was what you mean?

• Log Insight 3.0 integrated Load Balancer application

  So I have a cluster of Log Insight of three nodes and active integrated load balancer entered the IP address and domain FULL that my clients are pointing to. Everything is good so far.

  I'd like to understand how balance really works, i.e. If one of the nodes becomes unavailable id still wait to be able to ping the address of the ILB?  the behavouir can I see at the moment is when the master is down so is the address of the ILB, is - this planned?

  Yes if your IP ILB was linked to the master, when he went to bed, the ILB IP is supported by another node, which explains why the ingestion continues. Which means that you can ping the IP, but if you access the UI through the VIP it tries to go to the user interface of the master master having died the user interface is not available.

• DRS load balancing

  Hi people,

  Based on a few blogs now I understand now DRS load balancing based on active memory used on ESXi individual cluster nodes. Yesterday, we migrated VM to an ESXi node which pushed the memory consumed 99% but DRS seems never disturbed. I assumed that justification as the memory usage active only 11 GB of 160 GB.

  But my query is that what happens when I migrate several virtual machines to hosts that already shows alarm for 99% of the consumption of memory? If the DRS was examined only active use so why he throws alarms based on consumption, creating panic with the customer?

  Please have helped to understand the same.

  Thank you!

  Best regards

  Hari.

  Hi Hari,

  Good point.

  DRS behavior can be changed in order to study the memory consumed as well. U2c 5.5 and above, DRS considers the memory consumed as well. How much he believes is below specified KBs.

  Reference: https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2059868

  There is a KB linked to top Ko. Please throw in that as well.

  According to me, the alarm is little generic, it works even without the DRS, it seems, the alarm is designed in a way that it is not aware of this host belongs to active DRS cluster. DRS comes with the license of the company and alarm will work with basic licenses as well.

  are you suggesting that alarm should be aware of the hosts that are part of the active DRS cluster and give only the alarm according to the metric memory configured on DRS?

• vRA 7 load balancing with embedded active / standby vRO

  I have a vRA environment 7 with vRA HA devices that have integrated the vRO. I want vRO to an Active/Passive configuration, but when I do the load balancer declares the node with the vRO vRA Eve out of service.

  It seems that between 6.2 and 7 that VMware has changed the vRA health check to watch/vcac/service/api/health instead of/vcac/services/api/status. This new target now fail because it shows a service as in offline mode. While it is fairly clean and is probably better than the old health check, it kills the idea of an active configuration / standby vRO.

  Any ideas on how I can get around this? Thank you!!

  The problem that you are facing is that devices of internal load balancing is managed by HAProxy. I don't have * think * that it promotes health checks, this is why active/active approach for vRO. I suspect that if you want to use Active/passive, you will need to switch to external devices.

  Grant

• In transparent mode load balancer

  I tried to configure the load balancer to use transparent mode, so that the servers can see the IP address of the clients. Because I've selected nodes in the pool by using containers of VC and no IP address looks like the established load balancer connection using IPv4 and IPv6 servers and there is a limitation of current in transparent mode if double stack is used, I got this message:

  

  Once I removed the nodes using containers of VC and I used the IP address in the IPv4 format this message seemed no more. Is - what the expected behavior, no idea why this topic?

  Thank you.

  This is the behavior expected if your VM have IPv4 and IPv6 (VMs most have now both batteries activated).

  In this case, you must configure the IP VM @ manually.

  Dimitri

• Help setting up MIrage with load balancer

  I'm trying to configure Mirage to use a F5 load balancer. We simply indicated the Mirage server and received an alias DNS and VIP of the N/W team. I not configure Mirage specifically on an F5 before, should it work without the guard dog service settings configured (at least up to a point where the client can connect)? Or the watchdog service must be configured?

  For setting up F5, we recommend enabling SSL and using SSL session ID persistence, also increase value of persistence for a few hours delay Mirage can be long connected and you don't want them to bounce around. A checkup at the service of Mirage generally configure TCP/half opened on the F5 that verifies that the service of the Mirage port is open. Also enable connections unless the method of balance load.

• Without SSL on Apps Tier DNS load balancing

  Dear all,

  I am to set up load balancing DNS (non-shared file system) for my 2 nodes of level Apps. I followed the id of the note using Load-balance with Oracle E-Business Suite Release 12 [380489.1 ID] section 2.4, but the question is. my level Apps is not configured with SSL. Please tell me how to configure the DNS without SSL load balancing so that I can connect with http://

  
  

  Kind regards

  Aleem
  

  You can use the same instructions - replace https with http and 443 with your port number.

  HTH
  Srini

• Hyperic HA does support load balancing?

  Hello
  
  We try to implement architecture Hyperic high availability based on JBoss HA and distributed cache EHCache. Our version is 3.2.3 EE.
  The cluster configuration is quite simple with the HQ High Availability Guide.
  Thanks for this guide and the esayness of the configuration of the feature!
  
  My first question was, the agent, what server IP should I use? I put with node 1 and other agents with node 2
  = > the synchronization between nodes works very well.
  
  Therefore, I have some additional questions.
  
  1 / can I have a single point of entry for agents with some load balancing and failover capability?
  For example, if an agent is sending information with the Node1 and this node breaks down or the server is restarted for maintenance purposes, the agent information will be lost.
  
  2 / same question for the application, can I have a single point of entry URL with some load balancing and failover capability?
  
  Thanks for your help.
  
  Nicolas.
  
  Post edited by: njmn

  HQ Enterprise HA for failover, but do not support for load balancing.  You will need before your HQ multiple servers with a load (software/hardware) balancer which will relay agent to a live server.  The load balancer must be staged all agents in a single node of 'master '.  If this server is unavailable, another server will automatically become the master node and agents should be redirected to this node.

  Charles

• Two shared the Application layer without load balancer

  Hi all
  
  I created the system of Application of layer two without the load balancer.
  
  EBS: 12.1.3 DB: 11.2
  
  I have two URLS:
  
  https://XYZ:8002/OA_HTML/APPSLocalLogin.jsp
  
  https://ABC:8002/OA_HTML/APPSLocalLogin.jsp
  
  I run the auctoconfig on the last node abc. When I try to access the https://xyz:8002/OA_HTML/APPSLocalLogin.jsp, it automatically redirects to https://abc:8002/OA_HTML/APPSLocalLogin.jsp
  
  If autoconfig on xyz last execution, https://abc:8002/OA_HTML/APPSLocalLogin.jsp URL automatically redirects to https://xyz:8002/OA_HTML/APPSLocalLogin.jsp.
  
  I know in 11i is the behavior, but according to my understanding in R12 this shouldn't ' e be the case because another application works very well.
  
  Concerning
  Sourabh Gupta

  Please see (multiple Application nodes configured without Load Balancer Cause Login Redirection [1362885.1 ID]).

  Thank you
  Hussein