Re: [RFC 1/3] SQUASHME: Generalize the device cache so it can be used by all layouts

[Boaz Harrosh <bharrosh@panasas.com>]

On 09/13/2010 02:16 PM, Fred Isaman wrote:
> On Mon, Sep 13, 2010 at 2:50 AM, Boaz Harrosh <bharrosh@panasas.com> wrote:
>>> This doesn't look like it needs to be part of the initial single-layout
>>> submission. Why is it being labelled as a squashme?
>>>
>>> Trond
>>
>> No, this is a device_id cache per nfs-client it is part of the initial
>> getdeviceinfo call. Many "single-layout" may refer to the same device_id
>> therefor it is cached so the long getdeviceinfo call can be done only once.
>>
>> Cheers
>> Boaz
> 
> I agree.  I'll roll it into the next submission.
> 
> Fred

Thanks Fred. I would like if you can report that it actually works, as I have
not tested it.

BTW. For the objects I have decided to not keep a reference on the device_id
during the lseg life span, but just keep a reference to the underlying osd_device
so the device_id can be removed from the cache but still be used by a 
layout_seg/io_state. I have tested and it works and apparently it unmounts
cleanly without any reference leaks. Will post patches later once the new API
hits Benny's tree.

Thanks
Boaz