[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Handle-info] Failover handler server

Resolution requests are generally sent round-robin by handle resolvers to available mirrors, with some heuristics to avoid servers that are down or are slower to respond.  It is possible to specifically ask for authoritative resolution which is only sent to the primary.

It may be possible to configure a prefix with a special status message, that will be included on the hdl.handle.net error page when there are handle resolution errors.  If this seems relevant to you, please contact your MPA administrator (e.g. hdladmin@cnri.reston.va.us).

Robert

On Jul 17, 2018, at 3:32 AM, Jozef Misutka <misutka@ufal.mff.cuni.cz> wrote:



On 16 July 2018 at 19:10, Robert R Tupelo-Schneck <schneck@cnri.reston.va.us> wrote:
I don't see any straightforward way to do this.  Instead of adding one special failover handle server, why not add one normal mirror that actually returns the right information?

Having proper mirrors is not easy with > 20 independent centres each having their own handle server.

However, can you point me to the documentation that describes how and when *exactly* is the mirror used (e.g., round robin fashion or only when the primary is not available)? Section 1.2 from the Technical manual does not cover this topic in much detail.
 
Thanks,
Jozef


Robert

> On Jul 16, 2018, at 12:19 PM, Jozef Misutka <misutka@ufal.mff.cuni.cz> wrote:
> 
> Is it possible to configure the handle servers in such a way that there will be one failover handle server which will be contacted *only* if other handle servers are unreachable?
> 
> Currently, if a handle server is down the information returned is of zero value. The failover handle server could return a predefined URL with more information about the handle and e.g., whom to contact. 
> 
> (We try to address the "returned information of zero value" issue on the handle server itself where we monitor individual handles and if one is unreachable for a noticeable amount of time we change the URL to our failover one that contains basic information on what to do.)
> 
> Best,
> Jozef
> _______________________________________________
> Handle-Info mailing list
> Handle-Info@cnri.reston.va.us
> http://www.handle.net/mailman/listinfo/handle-info


_______________________________________________
Handle-Info mailing list
Handle-Info@cnri.reston.va.us
http://www.handle.net/mailman/listinfo/handle-info

_______________________________________________
Handle-Info mailing list
Handle-Info@cnri.reston.va.us
http://www.handle.net/mailman/listinfo/handle-info