> You can't do this the way you have set it up. The servers control which
> paths are stageable not the manager. If you want some managers to allow
> staging and some not, you need to run overlay clusters. That is you have two
> xrootd clusters that are identical but one cluster affiliates with a
> particular set of managers and has stageable paths and the other cluster
> (which sees the same space) affiliates with another set of managers but does
> not allow staging.
If I update the other manager stageable, will it work?
> As for the meta-manager question. The meta-manager is meant to be used for
> global clustering of distributed clusters. The manager/supervisor scheme is
> the only supported mechanism for clustering at a local level.
> ----- Original Message ----- From: "wen guan" <[log in to unmask]>
> To: "Andrew Hanushevsky" <[log in to unmask]>
> Cc: <[log in to unmask]>
> Sent: Tuesday, August 24, 2010 2:10 AM
> Subject: Re: Is it possible to use more than 64 servers in xrootd staging
>> Hi Andrew,
>> Thanks for the information. Can I use meta manager instead of
>> Below is my cfg. If I use atlas-bkp2 to read a file which is in
>> c170~c199, xrootd will try to stage in this file. But I don't want it
>> to stage in this file, I want it to read the file from c170~c199. What
>> do I need to update?
>> On Tue, Aug 24, 2010 at 4:00 AM, Andrew Hanushevsky <[log in to unmask]>
>>> Hi Wen,
>>> I'm not sure I completely understand the question. It is always possible
>>> have more than 64 servers as long as you have the supervisors to support
>>> them. Any or all of those servers can be designated as staging servers.
>>> the answer is yes assuming you are asking the question I think you are.
>>> ----- Original Message ----- From: "wen guan" <[log in to unmask]>
>>> To: <[log in to unmask]>
>>> Sent: Sunday, August 22, 2010 7:25 AM
>>> Subject: Is it possible to use more than 64 servers in xrootd staging
>>>> I tested to setup a xrootd staging pool using more 64 machines. It
>>>> seems xrootd tries to staging the file from backend pool if the file
>>>> is in supervisor. Is it possible to use more than 64 servers in xrootd
>>>> staging pool?