Lowest node number rule in 2 nodes RAC when interconnect failure

Is it true that when a interconnect failure occurs in a two node cluster, lowest node number rule? Assume both the nodes can access the voting disk(s).
If true, any documentation or white pager to address more details?
Thanks in advance.
Limin.

Is it true that when a interconnect failure occurs in a two node cluster, lowest node number rule? Well, I am not sure this could be true or not. Couple of times its been observed in our two node RAC Databases where each time one node evicted due to couple reasons (one of the reason was the time synchronization between the nodes) and surprisingly another time a node evicted without any valid reason too.
Jaffar

Similar Messages

RAC interconnect failure

We have set up a RAC DB and we are trying to understand the effect of interconnect failure. Our understanding is that when interconnect fails, the RAC will use the information in voting disk to decide which node will survive and which node will be shut down. We want to make sure that is what is going to happen. Interconnect failure will not result in both node being shutdown, correct?
Please point us to the right oracle doc about this isssue, if one exists. We were not able to find firm answer on the docs we have read so far.
Thanks.

Is it true that when a interconnect failure occurs in a two node cluster, lowest node number rule? Well, I am not sure this could be true or not. Couple of times its been observed in our two node RAC Databases where each time one node evicted due to couple reasons (one of the reason was the time synchronization between the nodes) and surprisingly another time a node evicted without any valid reason too.
Jaffar

Migrate Single Node 10g STreams to Multi Node RAC

We are currently running 10g STreams. One capture process enqueing LCR into one queue. Propagation job dequeues messages and propagates them to a Destination queue. Then our Apply process at Destination database dequeues and apply changes.
We have developed a single shell script STREAMSsetup.sh that does all the magic on setting up streams (capture, apply, propagator and the table queues.)
Now, we are in the process of designing our new Horizontal architecture with Oracle RAC. We envision it is going to be a Data Guard Migration from Single node to the RAC environment.
My question is, can we just let Data Guard takes care of that, expecting all Streams pieces will be in place within the RAC environment? OR do you recomment we remove all streams configuration and re-run our magic script STREAMSsetup.sh to set everything up again once migrated to the RAC environment. ???
I now Streams follows an Instance Owner rule when it comes to where Capture or Apply will be started or stoped.
I would appreciate if someone have had any experience with this type of migration provide any feedback.
Thanks,
Arqui

Yes, we are opening an SR. Everything is pointing into the direction of recreating Streams configuration from RAC/ASM environment once migrated, using the instance with the lowest Instance number in the Cluster. That instance will be registered as the Primary instance owner of the Buffer queue. If primary instance fails queue ownership is transferred automatically to the next instance in the cluster and capture is started automatically in the other node.
Thanks,
ARqui

How do i find out the node number/name from the weblogic API ?

Hi
          I have written an application component that writes to a file. When clustered, i need to make the filename correspond to the managed server / node # inside the cluster, as i cannot afford to let each jvm hence each node write to the same file. Therefore on startup i would like to interrogate the weblogic server and ask it the node # the application is deployed in, so that it can create a unique file name - BASED upon the node number.
          Any ideas anyone ?
          Thanks
          James

You may want to use WebLogic JMX api to do that. Using MBeanHome, you will get the name of the server on which your application gets invoked.
          Regards,
          Vishwas

Getting node number/name in a multi node installation

Is there any way to get the number/name of the node in an installation with multiple nodes (Netweaver ce 7.1 java).

Thank you very much for your reply.
It's currently not realized, but we are thinking about:
A Netweaver 7.1 CE installation, Java Stack, as a cluster with more than one server node.
What we want is to identify the server node (by number or name) the current process is performed.
If such a feature will be available only in NetWeaver 7.2, upgrade is possible.

Not Hier node levels we need Hier Node Descriptions

Hi. Guys.
I come accross with issue like we have 0prod_hier with different hier node levels and we need Hier
node descriptions in place of node levels after executing the report. So
I have changed Level 01 Level 02 Level 03 ......... with Desc1 Desc2 Desc3...... like that in Hier Maintenance
from backend side.
I could be able to see the descriptions in places of Level 1 level 2 Level3 ....
But problem is when I navigate the 0prod_hier into rows/columns it is going to short dump. Why
I am getting I am not understanding
and when I navigate other fields I am not getting problem. Only for Prod_hier, i am facing this
problem error is saying like this. and currently Prod_hier is in Free Characterstics.
Error analysis
A RAISE statement in the program "SAPLRSDU_PART" raised th
condition "TABLE_NOT_EXISTS".
Since the exception was not intercepted by a superior
program, processing was terminated.
Short description of exception condition:
Please help me out how to solve this.
thanks & Regards
Vijay

Hello Vijay,
the cause of this problem are the temporary hierarchy tables.
You may execute the report SAP_DROP_TMPTABLES (in se38) for the temporary hierarchy tables (= 02-tables) and
recheck the behaviour. Please generate the query after doing this. You can find more information regarding temporary tables
in note 449891.
best regards,
Jerome

Context Mapping: child-nodes of non-mapped parent nodes

I am somewhat curious about Context Mapping in WebDynpro.
Which parents nodes need to be mapped in order to map child nodes ? Does a child node has more than one parent node (e.g. grandparent - two steps above) ?
A.1 -
MAPPED----
>    B.1
- A1.1   NOT MAPPED
- A1.2   NOT MAPPED
        - A.1.2.1     MAPPED TO>    B.1.2.4
        - A.1.2.2   NOT MAPPED
- A1.3   NOT MAPPED
Is it sufficient that only one of these parent nodes (e.g. direct parent node not mapped, but parent node of this parent node is mapped) need to be mapped so that the child node can be mapped ?
Is this assumption true or not ?
The SAP library states:
"Conversely, child nodes of non-mapped parent nodes cannot be mapped, otherwise this would result in irresolvable conflicts at runtime with respect to the parent-child relation in the context and the mapping relation." (http://help.sap.com/saphelp_nw04/helpdata/de/51/a3384162316532e10000000a1550b0/content.htm)
This statement is not absolutely clear on this issue.

Let me put it this way
You have a node vehicle and you have a child node for that the car. The car node have parameters car1, car2.
Let the vehicle node have parameters veh1 and veh2
Then if you map vehicle node one to any other node (say in the comp. controller) you have the option of mapping its children i.e veh1,veh2 and carnode.
You can have that veh1 is mapped and veh2 is not mapped.
vehicle node(mapped)
veh1(may or may not be mapped)
veh2(may or may not be mapped)
car node(may or may not be mapped)
If instead you try to map only the car node the vehicle node will also get mapped automatically but not it's child parameters like veh1 and veh2.
vehicle node(mapped)
veh1(may or may not be mapped)
veh2(may or may not be mapped)
car nodemapped)
car1(may or may not be mapped)
car2(may or may not be mapped)
Hope this would help.
Do revert for further clarification
Regards
Noufal

FTREE_NODE: Find parent node which has no child nodes

Hi guys
Does anybody know how to find a parent node which has no child node(s)?
I couldn't find any build-in at the online help. Is there a trick to got this?
Thanks in advance
Remo

Hi Remo,
I misspoke, in my earlier post. The FTREE built-in has defined constants for the possible values that FTREE.NODE_STATE can have -- FTREE.COLLAPSED, FTREE.EXPANDED and FTREE.LEAF_NODE.
If you wish, rather than testing for a value of '0', you can instead test for a value of FTREE.LEAF_NODE. Both are functionally identical, but using the constants can improve the readability of your code, eliminating the need for additional comments.
Eric

ADF Tree setting focus back to parent node after deletion of child node

Hi,
Is there a way to get the focus back to the parent node (or rather any particular node) in a tree?
I have a use case where we need to get the focus back to the parent node after a child node is deleted.
Currently the focus is shifted to the next node in the tree, but the need is to get the focus shifted back to the parent node. Also the parent node should be re-invoked to populate to get the latest status after deletion of the child node.
Any help/pointers?
Thanks

Thanks for the reply Frank.
I saw the link http://sreevardhanadf.blogspot.in/2012/07/showing-next-row-as-current-row-after.html
However the issue is since I am using custom created tree using POJO tree item (composite object).
calling myTree.getWrappedData() doesn't gives me a handle to JUCtrlHierBinding and subsequent access to JUCtrlHierNodeBinding.
my program gives me data like -
List<MyTreeItem> treeData = (List<MyTreeItem>)treeModel.getWrappedData();
because my tree model is build using -
treeModel = new ChildPropertyTreeModel(items, "children");
where items is List of <MyTreeItem>
Hence I am unable to get a handle using -
List nodeParentList = nodeParent .getKeyPath();
I am programmatically able to invoke the parent node to get the fresh data, only issue is the focus/selection of that node is not happening
Is there a way around?
Thanks
Sachin

Casting a org.w3c.dom Node to a org.dom4j Node

Does anybody know how I can cast (convert) a w3c.dom Node to a dom4j Node?
If I just trying casting the w3c node "(org.dom4j.Node)" it throws an exception...

I have a org.w3c.dom.Node after calling HttpUnit's getDOMSubtree() method.
Now I would like to change that Node to an org.dom4j.Node so i can use dom4j's Node.valueOf() method on it.
Any idea how?

Action Script to add a parent node value to the top node

I need to add a new parent node value to a hierarchy for the top node. The action script parameters to add a node is "Add|Version|Hierarchy|Node|Parent Node|Leaf property". I am at a loss on what to do for the top node since there is no Parent Node above this top level, but the parameters require it. I have reviewed both the user manual and admin manual, but did not find any assistance. Can anybody help me?
Thanks!

You have the answer with you, don't get confused between Parent Node and Top Node in this context, while Adding a new node your Top Node will become your Parent Node.
"Add|Version|Hierarchy|Node|Parent Node|Leaf property"
If you still have confusion, simply do the below,
1. Right Click on your Top Node and Add a new Leaf or Limb Node manually
2. Once added, go back to Home -> Script
3. Select Source Type as Transaction Log and then select the newly added node and Load
4. In the bottom select Script -> Download -> Download as Action Script.
5. Open in a Notepad and leverage it to create your new script.

How to access data from node of other view into node of diff view?

Hi all,
 My requirement is that I have account defined in 'CLEAR_HEAD' component , view is 'CLEAR_HEAD/ClearAccountsEF'
context node is 'TARGETVALUE' and attribute is 'ACCOUNT1'. I need to access this account information in the same component but in another view 'CLEAR_HEAD/ClearAddressEL' and the context node is 'DEPLIST' and attribute is 'RADIO1'. I have one custom controller 'CLEAR_HEAD/CuCoHead' already defined in this component. I checked the above 'TARGETVALUE' node but it is not bind to any Custom controller. And I also checked the above pre-defined custom controller 'CLEAR_HEAD/CuCoHead' and the context node in this Custom controller is different from those of the above context nodes and the only one context node defined in this custom controller 'CLEAR_HEAD/CuCoHead' is 'LISTNODE'.
My quesitons are
1. How to bind the above context node(TARGETVALUE) to the custom controller. Can I bind it to the above existing custom controller 'CLEAR_HEAD/CuCoHead' , if so then how can I do since the above custom controller doesn't contain this context node(TARGETVALUE) and it contains only the context node 'LISTNODE'. If I can use the above existing Custom Controller then can I directly do binding by right clicking the context node(TARGETVALUE) -> Create binding. Then what should be the value in Target Contex Node?
2. Do I need to create another custom controller If so then can I directly follow the steps defined in this link '/people/vikash.krishna/blog/2009/12/28/crm-70-how-to--5d-custom-controller-and-binding-of-context-nodes' ? If so then what should be the Model node and Bol Entity name?
I appreciate your help..
Thanks,
Chinnu.

HI Chinnu, 
 
Approach 1: 
Bind the context node view1 TARGETVALUE with Component controller. 
and in View2 
data lr_comp_controller type ref to X----component controller 
lr_entity TYPE REF TO cl_crm_bol_entity. 
lr_comp_controller ?= me->comp_controller. 
 
lr_entity ?= lr_comp_controller->typed_context->Y->collection_wrapper->get_current( ). Y---CONTExt node in view1. 
 
Approach 2: 
Declare a static attribute in the View controller of first view. 
Read the attribute D1 value from DO_PREPARE_OUTPUT of first View and pass it to the Static sttribute. 
In your 2nd view u can read this Static Atribute and use it. 
Static Attribute u need to call as follows . Z********IMPL(Implemantation Class)=>ATTR1 
 
Approach 3: 
Creae a custom controller with the same context node used in view 1. 
Bound the custom controller context node to view 1 context node. 
 
Now in view 2, get the custome controller and read data from it. 

 
View1 context node: BTADMINH 

Custom controller: BT115QH_SLSQ/AdminH context node: BTADMINH_CUCO 

In view 1 context class (CTXT class), go to CREATE_BTADMINH method and append the below code to bind both the context nodes, 
<pre class="jive-pre"><code class="jive-code jive-java">* bind to custom controller
 owner->do_context_node_binding(
 iv_controller_type = CL_BSP_WD_CONTROLLER=>CO_TYPE_CUSTOM
 iv_name =
 'BT115QH_SLSQ/AdminH' "#EC NOTEXT
 iv_target_node_name = 'BTADMINH_CUCO'
 iv_node_2_bind = BTADMINH ).
Now in view 2 get the custom controller using the below code,
Data: lr_cuco type ref to <custom controller class>.
lr_cuco ?= me->get_custom_controller( controller_id = 'BT115QH_SLSQ/AdminH' ).
lr_entity = lr_cuco->typed_context->BTADMINHCUCO->collection_wrapper->get_current( ).
</code></pre> 

If you are just reading a attribute value from view1, then approach 2 shud b easy one. 
Approach 1 & 3, more or less the same, except we are using component controller vs custom controller. These approaches holds good when the data required to be accessed/updated from either views. 
 
Coming to your specific scenario, I think there is already a context node available in the compoenent controller TARGETBP, which can have the required data for you. Please look into that. 
 
Cheers, Satish 
Edited by: Satish Reddy Palyam on Aug 13, 2011 12:50 PM

Could not load or refresh node tree creation failed on node

Hi Experts
I created one user with MSS role and ERP COMM role.
It is working fine. We restarted the server for other purpose. Then surprisingly the MSS role was locked. I am unable to see the MSS tab in the top level navigation.
When i expand the workset in line manager -> Manager Self Service -> Worksets
I am getting the following error::
Could not load or refresh node tree creation failed on node:
pcd:portal_content/com.sap.pct/line_manager/com.sap.pct.erp.mss.bp_folder/com.sap.pct.erp.mss.worksets
I can be able to expand and priveiw the Pages and iviews, except the worksets
Please guide me, i am badly stucked here.
Regards,
Srinivas

Hi KRISHNA,
Check the below thread and notes mentioned in that
Re: Tree creation failed on node: pcd:portal_content
Koti Reddy

Would the CSS style set to a parent node be inherited to underlying nodes?

Hi,
Would the CSS style set to a parent node be inherited to underlying nodes?
With below code, style setting to parent(VBox) is not getting inherited to children(Text).
Style.java
package style;
import javafx.application.Application;
import javafx.scene.Group;
import javafx.scene.Scene;
import javafx.scene.layout.VBox;
import javafx.scene.paint.Color;
import javafx.scene.text.Text;
import javafx.stage.Stage;
public class Style extends Application {
 public static void main(String[] args) {
 Application.launch(Style.class, args);
 @Override
 public void start(Stage primaryStage) {
 primaryStage.setTitle("Hello World");
 Group root = new Group();
 Scene scene = new Scene(root, 300, 250, Color.LIGHTGREEN);
 scene.getStylesheets().add("/style/style1.css");
 // Create children
 Text text1 = new Text("Text1");
 Text text2 = new Text("Text2");
 // Create parent and set style
 VBox vBox = new VBox();
 vBox.getChildren().addAll(text1, text2);
 vBox.getStyleClass().add("font-normal");
 root.getChildren().add(vBox);
 primaryStage.setScene(scene);
 primaryStage.setVisible(true);
}Style1.css
.font-normal {
 -fx-font-size: 9pt;
 -fx-font-family: "Arial";
 -fx-fill: #FF0000;
}Edited by: Simosh on Jun 10, 2011 4:24 PM

Hello Simosh,
I know you are thinking of VBox as container so their children must have inherited the css of parent like as of HTML <div> pattern. But in JavaFx this is different every node has it's CSS implemented from the Parent of their extended class . Even if you add children inside VBox CSS has nothing to do with children. So to make children to use the css of VBox then you must call getStyleClass().add() explicitly for every children.
Either you call getStyleClass() for every node
or you can make a custom Container as on below
A sample getChildren() overriding of VBox which make VBox children to use their parent css.
Algorithm to make your custom VBox
class MyBox extends VBox{
 @Override public ObservableList<Node> getChildren(){
 //iterate children of super
 //now bind your children.styleclass with super.styleclass
 //end of iterate
 //return children
}Thanks.
Narayan

Moving active node resources to the inactive node

We have a 2 node , majority disk, Windows 2008R2 failover cluster.   The active node is currently showing all resources and services on node A.   Node B is the inactive node.    we are about to perform a DataCenter shutdown
test.   Our application manager would like us to make Node B inactive during the DataCenter Shutdown test but he believes we can make Node B active if we do it this way:
1. power down Node B (inactive node)
2. power down Node A (active node)
3. power up Node A (active node)
4. power up Node B (inactive node)
By doing this way, he believes Node B will automatically become the Active node.
I do not believe this is correct.   I believe it has to failover to get Node B active. So, my proposal is:
1. power down Node A (active node) This should cause a failover of resources and services to Node B
2. Once all resources and services are on Node B, power Node B down.   Now, both nodes are powered down and Node B is holding the cluster resources and services and is the Active node.
3. Power up Node B (now the new Active node)
4. Power up Node A ( the new inactive node)
Which is correct?
Thanks in advance

Earlier question on ownership ... In a two-node cluster, ownership does not mean much. In multi-node clusters, it is often used with the anti-affinity property to try to keep certain resources from running on the same node. For example, if you
have a four node host cluster running two VMs that are also in a cluster, you would most likely not want both of the SQL VMs to reside on the same node of the cluster. Using anti-affinity and preferred owners, you can pretty much ensure the two VMs never
end up on the same node. In a two node cluster, ownership doesn't mean much.
When SQL is installed in a cluster, SQL is installed on two or more nodes of the cluster. SQL is actually running on each node on which you have installed it. In your case, both nodes are running SQL. By default, if something happens to
the SQL instance running on one node, it will try to restart on the same node, i.e. it will not immediately failover. There is a retry count and a time limit set up on the resource. If the retry count is reached in the specified time limit, then
the resource, in your case the SQL instance, will fail over to the other node of the cluster.
In the Failover Cluster Manager console you will see a selection option for Cluster events. That will show you all the event log entries that have occurred on both nodes of your cluster relating to cluster events.
.:|:.:|:. tim

Lowest node number rule in 2 nodes RAC when interconnect failure

Similar Messages

Maybe you are looking for