Active Passive Vs Active Active

Hi All,
How much more effort it requires to failover or get up and ready with passive node, if Active Passive strategy is followed rather than Active Active just for saving on licenses.
Note : High availability is not a concern.
thanks,

Be clear what the policy means. RAC is normally considered an ACTIVE/ACTIVE solution since all nodes in RAC are actively processing. In an ACTIVE/PASSIVE setup by contrast one or more nodes are actively processing whilst others are standing by ready to take over the workload in the event of failure of the primary. Physical Standby, Failsafe and non-oracle clusters are often Active/Passive. You may also check if the policy comes from a sqlserver centric view of the world where you split the sqlserver databases between different nodes, so that although for any given sql database the solution is active passive, you don't 'waste' the resources of some nodes just waiting to take over, but rather run other databases on them.
Your description of ACTIVE/ACTIVE seems to rather correspond to a Stretch Cluster http://www.oracle.com/technology/deploy/availability/pdf/1126_Ray_WP.pdf has a good overview of what this means. I'd be surprised if the policy was really to put all databases on just about the most expensive hardware and software solution available, but I'm used to surprises in the world of corporate policies.
Niall
user526463 wrote:
My company was recently purchased. The new company has a "policy" about any High Availability database be an Active\Active configuration. We have in the past used active\passive RAC installs with each site have multiple active RAC nodes. I am interested in see what the cases are for a true Active\Active environment where there will be multiple RAC nodes in each "cluster" and multiple databases (disk) which are actively written to, then synchronized. (As apposed to a couple mutinode RAC setups with a primary\standby database)
Any white papers or best practice document references would be great. I'd hate to do this just case of some policy.
Thanks

Similar Messages

Install of Data Services as active/passive behind F5 load balancers

I currently have two RedHat 5.5 x86_64 hosts, SAP_DS1 and SAP_DS2, and 1 database, Oracle RAC 11g and 1 F5 LTM load balancer sitting in front of the two hosts. The F5 load balancer VIP is set up as active/passive (fail-over) for the two hosts.
Will I need 2 local databases 1 central database or can I accomplish this with one local database?
Please advise on the recommended method for installing BOBJ DATA SERVICES XI 3.2 on two Linux hosts in an active/passive design.
Thanks

You might have better luck asking this question in the (somewhat misnamed) [Data Integration and Data Quality Management|Data Services and Data Quality; forum:
This forum is dedicated to topics related to SAP BusinessObjects Data Services (Data Integrator, Data Quality Management, Text Data Processing), SAP BusinessObjects Information Steward (Metadata Management, Data Insight), SAP BusinessObjects Rapid Marts and SAP BusinessObjects Data Federator.
(emphasis added)
Regards,
Sean
P.S. to moderators: can you create a sticky post to this effect?

Lenovo DisplayPort to DVI adapter--is it active or passive??

part number 43N9159 is it a passive or active adapter??
Solved!
Go to Solution.

Lenovo doesn't sell Active Displayport to DVI-D adapter. So the answer to your question is it is passive.
If you need to get active one (Single link or Dual Link DVI-D) then you would need to find it from third party retailers or OEM. Try ebay.
Regards,
Jin Li
May this year, be the year of 'DO'!
I am a volunteer, and not a paid staff of Lenovo or Microsoft

Switch over from active to passive in the cluster

If online EMS Sun Server goes off line , the client looses the connectivity with EMS till the IP address of the other online server is updated in client PC.
When the servers switch over from active to passive in the cluster, we have to change the server IP settings at the client PC manually. Since both the active as well as the passive servers will be pinging, we are not in a position to identify the active/passive servers, so automatic server setting is not possible at the client.
How to find out the solution for which server is active when the swich over is happend.

Hi Kristoffer,
remember to click on enable schedule when you schedule a package.
To know if a scheduled package is ok you can see first if present in the tblschedule of your appset and also if present in the windows task schedulerif with status "ready".
After you just have to check the view status if you found the package execution completed you're sure it's running. " version at the same time can cause issues also with different packages better to schedule with a sufficient delay to be sure that the first has ended, in your case 1 package is ok.
I would delete the 2 packages from bpc, checking also the windows scheduler to be sure that no packages are running and rebuild one from scratch setting repeat every 4 hours and checking "enable schedule".
Regards
Roberto

How to Configure Active-Passive oracle cluster in Windows 2008 R2 64bit Server.

How to Configure Active-Passive oracle cluster in Windows 2008 R2 64bit Server With Oracle 11g R2.
How many database will play in this role.
Best,

hello
I was going through your post and i am also doing the same thing here at our organisation for Oracle 10g R2
Can you pls send me any docs u r having for configuration of Oracle in windows clusters .
And, can you pls elaborate on this point
e)Create Oracle Service with the same name in the 2nd node and copy all the files like spfile,tnsnames.ora,listener.ora,password file to Node2.
Pls send me the details at [email protected] or you can contact me at 08054641476.
Thanks in advance.

Can you Setup a Active Passive DFS namespace in the same site

Hi,
I would like to use DFS for profile replication. I know MS doesn't recommend it, but was wondering if you could setup DFS namespace in an Active/ passive contests in the same AD site. I think or what I can understand of DFS is that if the Fileserver is in
the same site it would select it randomly and you cant specify which server is the primary.
Basically what I want is to have users only access File server 1, but if this file server 1 goes down to be redirected to file server 2.
Thanks,
Shaun

You could have a read-only Replica but that really doesn't help you in terms of the automatic redirection thing. You could also set up a 2-node replication group and disable the referral for the folder(s) on one node like below:
However the "failover" would be manually Enabling, which I doubt fits the bill.
Otherwise, I'm not aware of any configuration of DFSR that allows you to do what you're talking about (maybe someone else will correct me.) HOWEVER, I suspect a design that fits your use case may be to create a DFS namespace that runs
on a file server cluster. That way if Node1 goes down, the cluster fails over, and the data continues to be available on Node2 when accessed via DFS.
If interested, there are a couple ways to approach DFS in combination with a cluster.
Configuration1: The cluster is configured to host the namespace as well as the folder targets (where the actual data is). CAVEAT - The cluster can only be configured as a standalone cluster.
Configuration2: Create a domain DFS namespace that resides on other servers.
\\domain\DFSRoot for example. Then point each of the leaves/folder targets to clustered shares. So you'd create a subfolder called "Profiles" that points to the clustered share that houses the profiles.
Just an idea. Hope this helps.

Error with concurrent users- Activation Passivation Bind variable ?

I have a programmatic view object based on procedure call that returns a ref cursor.
Application Module has a function that exposes get Method for this View object using client interface.
Everything works good until many users call the same Method ,for many concurrent users i got the below error intermittently :
JBO-27122: SQL error during statement preparation.
Attempt to set a parameter name that does not occur in the SQL: bussId
For 1 or 2 users this works perfectly fine.
bussId is a bind variable defined at View object of type Array.
I understand this should be because of activation/passivation as many users calling same function comes to picture.
To solve this I overloaded activateState/passivateState in View implementation for the bind var.
But i don't see those methods getting invoked ever.
I also enabled connection pool/AM pool and kept the default pool values high enough for concurrent users.
Nothing works .
Appreciate any help in this regard.
-Sam

Hi,
See
http://antonis-antoniou.blogspot.com/2012/02/ensuring-that-your-adf-application-is.html

Non-domain user authentication against SSAS on Active/Passive Cluster

Hello,
We have an Active/Passive SQL Server setup (DB1 & DB2 Servers) connected to a cluster for SQL & SSAS. I have a web server not on the same domain that I am trying to authenticate with SSAS. This works OK if I set the website to impersonate
myUser and I add local account myUser as an Admin on SSAS for the active server (DB1). But when this fails over to DB2 then it fails to authenticate. SSAS won't allow us to add myUser as an admin for local accounts on both DB1 & DB2 as it errors
adding the second one. Could anyone advise how such a scenario should be approached?
We have tried creating a domain user too which DB1 & DB2 can of course both share but I don't think the web server can impersonate this with being not part of the domain.
Thanks.

Hi Jcorker,
According to your description, you need to access the SQL Serve Analysis Services database which is configured as cluster for SQL & SSAS from another domain, right?
In SSAS we can use the solution below achieve the requirement.
1.Create new domain account and impersonate the web site with that.
2.Create local user account on the analysis service with same exact username/password as like domain account created in the previous step.
However, you cannot create a local account with the same name on both servers. I have tested it on my local environemnt, we can create the same local account with the same name on both servers. In your scenario, if DB1 and DB2 on different server, you can
create a local account with the same name on both servers. Please post the detail errors, so that we can make further analysis.
Besides, SSAS only allows users of the same domain or trusted domains and it does not allow users from any domain except from these two. You can configure the trust relationship between the domains.
http://technet.microsoft.com/en-us/library/cc961481.aspx
Regards,
Charlie Liao
If you have any feedback on our support, please click
here.
Charlie Liao
TechNet Community Support

How to : Active/Passive and Active/Active Enterprise manager R3 (10.2.0.3)

Soon I will implement the above named version of EMR3 on a AIX 5.3.3 platform. I have a couple of questions about active/passive and active/active implementations. These are the questions I have:
Basic install:
Implementing 10.2.0.2 and upgrading to 10.2.0.3 works fine on host 1234 (already done that).
Does anyone have experience with an implementation on host 1234', in other words, a virtual host (different IP-numbers on the same ethernet adapter)?
Active/passive
Lets assume that the implementation on server 1234' works fine, how can I realise a cold failover situation? On AIX you can move/transport the volume group to the failover server, activate the volume group, mount your filesystems and the application is running again. This is the theory but ....
Active/Active
If this 'basic' install is done, can someone briefly tell me how to migrate the OMS repository to a RAC implementation? Is this possible or do I need to setup the repository for the OMS at the start of the implementation. Has anyone experience with using virtual hostnames, RAC and EMR3.
Can anyone tell me where I can find some good documentation instead of reading all these thousands of pages of Oracle's guides.
Help would be appreciated.
Message was edited by:
Jeroen Udo

Soon I will implement the above named version of EMR3 on a AIX 5.3.3 platform. I have a couple of questions about active/passive and active/active implementations. These are the questions I have:
Basic install:
Implementing 10.2.0.2 and upgrading to 10.2.0.3 works fine on host 1234 (already done that).
Does anyone have experience with an implementation on host 1234', in other words, a virtual host (different IP-numbers on the same ethernet adapter)?
Active/passive
Lets assume that the implementation on server 1234' works fine, how can I realise a cold failover situation? On AIX you can move/transport the volume group to the failover server, activate the volume group, mount your filesystems and the application is running again. This is the theory but ....
Active/Active
If this 'basic' install is done, can someone briefly tell me how to migrate the OMS repository to a RAC implementation? Is this possible or do I need to setup the repository for the OMS at the start of the implementation. Has anyone experience with using virtual hostnames, RAC and EMR3.
Can anyone tell me where I can find some good documentation instead of reading all these thousands of pages of Oracle's guides.
Help would be appreciated.
Message was edited by:
Jeroen Udo

Passive or active VGA-splitter for Mini?

I would like to mirror my Mini display with VGA-input to a 32 inch LCD-TV with VGA input. Should I use a passive or an active VGA-splitter?
mac mini Mac OS X (10.4.8)

Active splitter devices appear to the computer as a
monitor, and then transmit the graphical data to the
displays connected to them.
I think that point needs to be repeated and reinforced!
Since the mini has a display detect line in the video connector, it receives signals from compatible devices that signal what they are and helps the system determine what resolution and refresh rate to set. An active splitter can feed appropriate data back to the mini, while a passive one can only relay what it receives from the display. If both displays are the same, that wouldn't be a problem, but if they are different, the mini will receive a detect signal that at best will suit one display but may not match the other, since it does not know there are two displays connected.

A few ADF BC state passivation/activation problems

Our development team has encountered and researched a few ADF BC passivation/activation related bugs that are very annoying. Some of them have been fixed in the new JDeveloper 11gR1 (11.1.1.1.0) but other ones are still active in the last production release. As we consider some of these bugs as very important we decided to report them in this forum:
*1) New (modified) values of read-only SQL-derived or transient attributes are not activated if they have been modified in ViewObject's cache and the corresponding changes have not been posted to DB yet*
The attribute values are included into the passivated state but the activation process does not try to activate a modified value if the method <tt>ViewRowImpl.isAttributeUpdateable(int)</tt> returns <tt>false</tt> for the corresponding row attribute. There are many business cases when some ViewObject attributes have to be modified despite the fact they are read-only (or conditionally read-only depending on the values of another attributes).
*2) Transient values (values of SQL-derived or transient attributes) that have been modified in ViewObject's cache to <tt>null</tt> are not cleared on state activation*
This is true even for updateable attributes. The problem exists because the passivation process does not write into the passivated state anything about attributes having <tt>null</tt> values. As a consequence the activation process does not try to modify values of the corresponding attributes during the activation and they keep the value fetched from the ViewObject's query.
*3) State activation process does not restore ViewCriteria row names*
As a result the query components (<af:query>, <af:quickQuery> and table filters) do not work correctly after state passivation/activation because they do not clear eventual previously applied filter condition as they do that by searching ViewCriteria rows having particular names. In this way the new condition is appended instead of overriding eventual previous one.

Hi Maxa,
I am the thread owner. Sorry for the delay, but my forum account was destroyed during the transition to the new Oracle's SSO, so I had to create a new account and I have missed your posting.
The table filter problem was resolved in the new JDev 11.1.1.2. The bug number is "Bug 8650239 - FILTERING CRITERIA ARE APPENDED WHEN AM POOLING IS DISABLED".
If you have to use previous JDev 11g version then the you have to patch the class oracle.jbo.ViewCriteria in the archives adfm.jar and oracle.adf.model_11.1.1.ear as follows (look at the bold lines):
{font:Courier}
public void passivate(XMLDocument doc, XMLElement node) {
ViewCriteriaRow r = (ViewCriteriaRow)row;
for(int j = 0; j < r.getAttributeCount(); j++) {
if(rowNode == null) {
rowNode = doc.createElement("Row");
((XMLElement)rowNode).setAttribute("uc", r.isUpperColumns() ? "1" : "0");
((XMLElement)rowNode).setAttribute("cj", (new StringBuilder()).append("").append(r.getConjunction()).toString());
*((XMLElement)rowNode).setAttribute("n", r.getName() == null ? "" : r.getName());*
public void activate(XMLElement node) {
for(Node rowNode = node.getFirstChild(); rowNode != null;) {
ViewCriteriaRow vcr = createViewCriteriaRow();
s = ((XMLElement)rowNode).getAttribute("uc");
if(s != null && s.length() > 0)
vcr.setUpperColumns(Integer.valueOf(s).intValue() == 1);
s = ((XMLElement)rowNode).getAttribute("cj");
if(s != null && s.length() > 0)
vcr.setConjunction(Integer.valueOf(s).intValue());
String name = ((XMLElement)rowNode).getAttribute("n");
if(name != null && name.length() > 0)
vcr.setName(name);
{font}
You have to decompile the class, add the bold lines, compile the class and repack it into the specified archives. I am affraid it is not quite legal to patch Oracle's classes but I do not have other solution.

Oracle 10g Database Installation In Active-Passive Mode.

Good Afternoon !!!!
We are installing Oracle 10g in Active-Passive Mode on HP UX-11.3 , with ASM. Can someone help me with step wise installation procedure for the same.
Many Thank's
Rajeev.

Check on the release notes for your platform
This is for x86-64 Linux.. read it all, then you'll find the "6 Documentation Corrections and Additions" section you'll find the RPMs that you need
http://download.oracle.com/docs/cd/B19306_01/relnotes.102/b15666/toc.htm
then go to the official doc installation of 10gR2 on x86-64 Linux
http://download.oracle.com/docs/cd/B19306_01/install.102/b15667/toc.htm
Yum is a package management tool. This is what you'll be using to install the RPMs. But first you'll have to setup the Yum repository, usually what I do is stage the DVD media on the server then make it as a repository. If ever there's a specific (higher) version that I need then I just pull it on the RedHat Network.
There are a lot of HOWTOs about Yum. You may read on this link http://docs.fedoraproject.org/yum/en/
- Karl Arao
http://karlarao.wordpress.com/

About passive and active

please tell me about passive mode and active mode of an adapter?

Hi
The definition of the adapter metadata should contain a global attribute Adapter Status, which activates or deactivates processing of the specific communication channel.
The Active / Inactive status of adapter indicates the status of message exchange.
You could either change this status from communication channel in ID or by RWB (Runtime workbench)
Refer below link for it
http://help.sap.com/saphelp_nw04/helpdata/en/44/19261d528e1a77e10000000a1553f6/frameset.htm
/people/sap.user72/blog/2006/06/04/starting-stopping-of-communication-channel-from-rwb
http://help.sap.com/saphelp_nw04/helpdata/en/9b/f12541c3d3f323e10000000a155106/frameset.htm
All adapters displayed have a red, yellow, green or grey status. The basic status of non-registered adapters is grey. Adapters that can only be monitored with the communication channel monitor also have a grey status.
The following list shows the possible statuses listed in their hierarchical order from top to bottom.
The following statuses are possible:
●      Green (): Channel is started and functioning without errors.
●      Red (): Channel is started but errors have occurred.
●      Yellow (): Channel is started, but inactive.
●      Gray (): Channel is stopped
●      Not available (n/a): Channel is not registered; no information available.
Channels with a yellow status () are started, but are inactive on all cluster nodes of a server cluster. This means that a channel is not processing any messages, even though it is started.
A channel has this status in the following cases:
■       The channel is configured as inactive in the PCK (it can also only be reactivated in the PCK).
■       The channel is set to inactive by the respective adapter (some adapters can do this and can also change this status if necessary).
Thanks
Swarup
Edited by: Swarup Sawant on Feb 11, 2008 8:16 PM

Cluster: automatic push of software from Active to passive node?

(Newbie) Setup: Two node cluster, active\passive.
After installing software on the active node will it (or any portion, e.g. config files) be automatically pushed to the passive node and kept in sync? Or am I required to install the same software on the passive node myself?
TIA,
edm2

To make sure your application can run from both cluster nodes, the executable should be installed on the same path on both nodes. This ensures that if the generic application fails over, the file upon which it depends are available on the
other cluster nodes.
Disclaimer: This posting is provided AS IS with no warranties or guarantees and confers no rights.

Active - Passive Cluster

Hey, I have to manage an active passive cluster on SLES10 SP4. Oracle 10.2.0.4
All oracle binaries are on a DRBD Device on mountpoint /oradata.
This mountpoint includes the binaries and all datafiles.
I havent installed these systems. At one stage on node died and got reinstalled by a systemadministrator.
After a switch from the running node to the standby node, the cluster is not starting.
Jan 3 14:36:01 a logger: Cluster Ready Services waiting on dependencies. Diagnostics in /tmp/crsctl.6114.
Jan 3 14:36:14 a logger: Waiting for Oracle CSS service to be available before starting
Jan 3 14:36:14 a logger: ASM instance +ASM. Wait 2.
After issung $ORACLE_HOME/bin/localconfig add the cluster is starting, the asm instance is coming up, and also the database instance is starting.
Shortly after that the db instance dies. Alertlog is showing something like recursive sql error.
Looks like writing to the datafiles is not working.
In asmcmd the diskgroups are showing DISMOUNTED (show MOUNTED before)
so what to do ?
a) Why is it necessary to issue "localconfig add" to make the cluster starting ?
b) What could be a reason for not being able to write to the database ?
c) How would you guys install an active/passive cluster ?
c1) Install OS on both nodes
c2) Install oracle binaries
c2.1) on both nodes in the local files system and do the mapping to the drbd afterwards?
c2.2) only on one node and just switch the mountpoint ?
Chris

Hi,
first of all please clarify, if you have Oracle clusterware installed, or just the local configuration for local ASM.
There is a big difference between the 2, and also using DRBD has nothing to do with Oracle clusterware (and is not supported by Oracle).
a.) Localconfig may be necessary for a local ASM installation in 10.2. Main reason is, that the CSS service configuraton includes the hostname (which changed if you switched nodes). Hence may not be able to start ASM. With localconfig you are fixing this issue.
Very important: It is not a good idea to run this is a cluster configuration, since it will break things (a lot of things).
b.) If the ASM diskgroup got dismouned, then the database cannot write. Hence your problem. Why the diskgroup dismounted, you should find in the ASM alert.log
c.) I would go for RAC One Node. Which has this all preconfigured. So it makes life easier. Search for "RAC One Node" on OTN for more information. At least I would take a manual cluster, and do mirroring by ASM, and not DRBD....
c1.) Yes install Oracle cluster (or Grid infrastucture in 11.2) on both nodes. Like a RAC.
c2) Install database software on both nodes locally.
c2.1) No. No DRDB.
c2.2) Both nodes. Same installation.
Regards
Sebastian

SAP PI 7.1 High Availability, Active-Active or Active-Passive

We are planning go to HA system for PI 7.11
I've few doubts/questions regarding this.
Can we make Active-Active clustering or Active-Passive? Or itu2019s our choice?
If it is our choice, then is at Configuration level (Application level)? or hardware level?
What are additional steps we need to take for Active-Active clustering? (Which we are looking for)
What is recommended by SAP?
Any Pros & Cons.
Please reply me if you have any information regarding this, any information regards to this is greatly appreciated
Thanks
Sagar

Hi Sagar,
Please use Active-Passive for your SAP PI 7.1 High Availability system.
For more details, please refer to the sdn link:
High Availability
->High Availability FAQ
->What is an "active-active cluster"?
An active-active cluster consists of two independent server nodes at a minimum. The workload within a cluster resource is shared between the server nodes. If a cluster node crashes the processes are resumed by the remaining cluster nodes. An active-active cluster configuration means that a cluster resource is active on all cluster nodes. The aim of an active-active cluster is not only to provide high availability system but to distribute the workload between the cluster nodes. Applications with a very high workload like databases benefit from an active-active setup. Due to the SAP Central Services Instance is a light component an active-active setup does not make any sense. Therefore the SCS is implemented as an active-passive cluster resource.
Best Regards
Julia

Active Passive Vs Active Active

Similar Messages

Maybe you are looking for