Hi jeff

This mirrors my requirements exactly. We did think about a client-side but 
discarded it  for the same reasons. Why have 100 clients indexing when you can 
just have 1 or 2 servers

Very disappointing :-(

Phil Bartlett


Sent from my iPhone

> On 20 Aug 2014, at 18:37, Jeffrey Altman <[email protected]> wrote:
> 
>> On 8/20/2014 1:08 PM, Phil Bartlett wrote:
>> Hi
>> 
>> We have just installed OpenAFS across several sites. We would like to 
>> install somekind of centralised "search engine" that can index AFS realms. 
>> Can anyone recommend one?
>> 
>> Phil Bartlett
> 
> There is no good indexing/search engine for AFS cells.
> 
> My requirements are:
> 
> 1. indexing by the file servers not the clients
> 
> 2. storage of the index on a per volume basis.  This is
>   required so that the index is correct for each readonly
>   and backup volume clone.
> 
> 3. a distributed search service that spans all file servers
>   so that a request issued to one file server is serviced
>   by all of the file servers that contain relevant volumes
>   without duplication of effort.
> 
> 4. result sets only contain matching items that are accessible
>   to the authenticated entity issuing the query.
> 
> 5. integration with client OS search libraries
> 
> Wish list items include:
> 
> 1. The ability to issue persistent queries that use the
>   AFS callback mechanism to notify clients when the
>   result set has changed.
> 
> 2. Automatic generation of thumbnails and other commonly
>   accessed metadata on the file server.
> 
> I want to implement these capabilities in YFS but I suspect it is going
> to be quite a while before that happens.
> 
> In the meantime, a client side indexing solution can be used for a
> single platform.  The downsides are the network traffic and the view is
> only applicable to a single user entity.
> 
> Jeffrey Altman
> 
> 
> 
_______________________________________________
OpenAFS-info mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-info

Reply via email to