Archived
This forum has been archived. Please start a new discussion on GitHub.
redundancy question
Is there any support to set up redundancy in ice? The situation I would like to set up is to pass in to the code a list of host/port combinations and pass that to the ice runtime so that when a proxy is requested the ice runtime will try all the endpoints before throwing an exception. (i.e. I have multiple redundant services which are running on distinct hosts and I would like the client to try all of them before giving up.)
In a similar vein, is there a way to set up something akin to a cluster of locators so that if indirect binding is used and the 'primary' locator service host is down that another locator host will be accessed by the client so that there would not be a single point of failure?
Thanks again,
Mayer
In a similar vein, is there a way to set up something akin to a cluster of locators so that if indirect binding is used and the 'primary' locator service host is down that another locator host will be accessed by the client so that there would not be a single point of failure?
Thanks again,
Mayer
0
Comments
-
Hi,
You can specify multiple endpoints in your proxy. For example:
MyObject:tcp -h host1 -p 10000:tcp -h host2 -p 10000
When an attempt is made to establish a connection using this proxy, Ice will select an random endpoint from those configured. In the event of an error Ice will try each endpoint before reporting failure to connect back to your application.
Regards,
Dwayne0 -
mcrystal wrote:In a similar vein, is there a way to set up something akin to a cluster of locators so that if indirect binding is used and the 'primary' locator service host is down that another locator host will be accessed by the client so that there would not be a single point of failure?
No, it's currently not possible to replicate the locator. If you have a commercial need for this feature, it's certainly something that could be implemented. Please contact us at info@zeroc.com if that's the case!
Benoit.0