>From Soumya :
Soumya has uploaded a new change for review.
https://review.gerrithub.io/295446
Change subject: Add gcc-c++ has Required package
..
Add gcc-c++ has Required package
With latest C++ compile changes, it is good
On 09/21/2016 11:09 PM, Soumya Koduri wrote:
> and another one (I lost the core) , but it was
> mdcache_lru_get->mdcache_lru_clean -> fsal_close()->close() . In
> FSAL_GLUSTER()->file_close(), below assert was hit -
>
> assert(obj_hdl->type == REGULAR_FILE);
>
> The obj_hdl->type was a large numbe
>From :
jlay...@redhat.com has uploaded a new change for review.
https://review.gerrithub.io/295458
Change subject: Have posix2fsal_attributes use the posix standard for time
fields by default
..
Have posix2fsal_attributes u
>From :
jlay...@redhat.com has uploaded a new change for review.
https://review.gerrithub.io/295457
Change subject: ceph: fix several bugs with the USE_FSAL_CEPH_LL_LOOKUP_ROOT
..
ceph: fix several bugs with the USE_FSAL_CEPH
>From Daniel Gryniewicz :
Daniel Gryniewicz has uploaded a new change for review.
https://review.gerrithub.io/295524
Change subject: Re-add owner validation cache optimization
..
Re-add owner validation cache optimization
2.
Branch next
Tag:V2.4.0
Release Highlights
* ntirpc V1.4.1
* cache_inode replaced with stackable FSAL_MDCACHE
* support_ex FSAL API extensions to allow associating file descriptors
or other FSAL specific information with state_t objects
* abort() on ENOMEM rather than attempt to continue
*
Crud, while working on the documentation for support_ex, just came across a
method other folks implementing support_ex missed...
There is the possibility that two open2() calls to open by name will race,
resulting in two fsal_obj_handle's being created. When the MDCACHE attempts
to cache the 2nd e
submit a change; I dont know if rgw needs share reservations, can we refuse
them?
Matt
- "Frank Filz" wrote:
> Crud, while working on the documentation for support_ex, just came
> across a
> method other folks implementing support_ex missed...
>
> There is the possibility that two open2(
I have already discovered a bad problem in V2.4.0...
Oh, and there's also the performance degradation from doing permission
checks on read and write all the time.
With all of this in mind, we will definitely focus on immediate fixes for
2.4.0 in the next few weeks, and cut 2.4.1 shortly after Bak
> submit a change; I dont know if rgw needs share reservations, can we refuse
> them?
You can't refuse them... Any NFS v4 OPEN is a share reservation. From
non-Windows clients, the share reservation may not have deny modes, but the
access modes must be tracked so that conflicting Windows share
I was testing Ganesha NFS V2.4.0 using the POSIX FSAL on a CephFS file
system, using the Ceph FUSE interface, and I am seeing ganesha.nfsd
killed by an ABRT signal. I was using the same ganesha.conf file that
worked with V2.3.2 and CephFS. Here is the gdb backtrace:
[Thread debugging using libth
11 matches
Mail list logo
