• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 806
  • Last Modified:

Limited distance between 2 nodes in a cluster using Oracle RAC 11gR2

Hi,
My company are buying a multimedia content management system. We want to deploying it on 2 sites (their distance = 1500km) and want to sync database between them. Each site is a complete system that can work without other site. The system requirement for database on this system is Oracle RAC 11gR2. I find some info on google which said that 2 nodes in a cluster must be < 100 km but i can't verify this.
If this info is true, is there other solution else to synchronize database between 2 or more very far sites ?
Oh, the most important thing i want to know is limited distace between 2 nodes.
Thx
0
dash1986
Asked:
dash1986
3 Solutions
 
slightwv (䄆 Netminder) Commented:
Since RAC shares a disk sub-system you would need some disk management what could span this distance in a pretty efficient manner.  That is likely the limiting factor.  It also needs a quick heartbeat interconnect.

As far as a remote sync I would look at streams replication or if one of the remote databases can be read-only, take a look at active dataguard.
0
 
jgiordanoCommented:
Oracle RAC is very sensitive, From experience if the cluster has issues with the interconnect they will recommend that all nodes connect to the same switch. You may have failover interconnects going to a different switch but all primaries should be on the same switch and all failover interconnects should be on the same switch. Also they recommend a 1g minimum connection.
0
 
Chakravarthi AyyalaDatabase AdministratorCommented:
Oh, the most important thing i want to know is limited distace between 2 nodes.
  => One thing i can recommend is, make sure that ping time between/among the RAC nodes is less than 25 milliseconds.  Usually, in all RAC implementations, we see this to be less than a milli second.  :)
Hope that answers your question about the distance.

Another piece of note:
Despite the all RAC nodes being in the same data center, many times we had to increase the CSS timeout from its default 60 seconds to avoid node evictions.
As jgiordano said, RAC is very sensitive.


0

Featured Post

Receive 1:1 tech help

Solve your biggest tech problems alongside global tech experts with 1:1 help.

Tackle projects and never again get stuck behind a technical roadblock.
Join Now