I stopped the rgw3 instance that is running on the same server (I am not
sure anymore if I can run radosgw and nfs-ganesha with the same
client.rgw3), started ganesha with this config. I am correct to
understand that the ganesha-rgw is not using the http(s) interface and
directly connects to ceph with librados (because that is why we have the
RGW config section)?
RGW {
cluster = "ceph";
name = "client.rgw3";
ceph_conf = "/etc/ceph/ceph.conf";
}
EXPORT {
Export_ID=301;
Path = "test:test3";
Pseudo = "/rgwtester";
FSAL { Name = RGW; User_Id = "test$tester1"; Access_Key_Id
="xxxxxx"; Secret_Access_Key = "xxxxxx"; }
Disable_ACL = TRUE;
CLIENT { Clients = 192.168.10.0/24; access_type = "RO"; }
}
Logs of ganesha related to test3 bucket
[@]# tail -1000 /var/log/messages | grep ganesh | grep test3
Apr 30 22:24:55 c01 nfs-ganesha[604956]: [Admin] mdcache_exp_release
:FSAL :INFO :Releasing RGW export for test:test3
Apr 30 22:24:55 c01 nfs-ganesha[604956]: [Admin] mdcache_exp_release
:FSAL :INFO :Releasing RGW export for test:test3
Apr 30 22:27:07 c01 nfs-ganesha[644097]: [main] export_commit_common
:CONFIG :INFO :Export 301 created at pseudo (/rgwtester) with path
(test:test3) and tag ((null)) perms (options=03303002 ,
, , , , , ,
)
Apr 30 22:27:07 c01 nfs-ganesha[644097]: [main] export_commit_common
:CONFIG :INFO :Export 301 created at pseudo (/rgwtester) with path
(test:test3) and tag ((null)) perms (options=03303002 ,
, , , , , ,
)
[@]# showmount -e nfs
Export list for nfs:
/ram 192.168.x.0
/test 192.168.x.0
/rgwtester 192.168.x.0
[@ mnt]# mount nfs:/rgwtester /mnt/nfstest
[@ mnt]# mount
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
tmpfs on /dev/shm type tmpfs (rw)
none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)
...
nfs:/test on /home/users/robbert/nfs-test type nfs4
(rw,nodev,relatime,vers=4,intr,timeo=600,nolock,retrans=2,proto=tcp,naml
en=255,hard,addr=
nfs:/rgwtester on /mnt/nfstest type nfs
(rw,vers=4,addr=192.168.x.x,clientaddr=192.168.x.x)
[@ mnt]# ls -l /mnt/nfstest/
ls: reading directory /mnt/nfstest/: Remote I/O error
total 0
-----Original Message-----
From: Daniel Gryniewicz [mailto:dang@redhat.com]
Sent: maandag 30 april 2018 20:06
To: Marc Roos; support
Cc: Matt Benjamin
Subject: Re: [Nfs-ganesha-support] Re: Nfs-ganesha rgw config for multi
tenancy rgw users
Okay, so on consultation, we think that maybe you should be mounting
test:test3 (ie, <tenant:bucket>) with no '/'. I can't tell from the
mail thread if you tried that; can you confirm?
(I have no personal experience with multi-tenant, so I didn't recognize
it as such).
In the mean time, we'll try to reproduce here and see what we can see.
Daniel
On 04/30/2018 09:28 AM, Daniel Gryniewicz wrote:
All of that looks fine, I think. I'm out of my depth here, so I
think
Matt needs to chime in here.
Matt?
On 04/30/2018 04:06 AM, Marc Roos wrote:
> Could it not be, because of a configuration issue in ganesha.conf or
> the version I am using? I can hardly imagine that I am the only one
> running a multitenant rgw config with ganesha-rgw.
>
> nfs-ganesha-xfs-2.5.5-.el7.x86_64
> nfs-ganesha-2.5.5-.el7.x86_64
> nfs-ganesha-rgw-2.5.5-.el7.x86_64
> nfs-ganesha-ceph-2.5.5-.el7.x86_64
> nfs-ganesha-vfs-2.5.5-.el7.x86_64
> nfs-ganesha-mem-2.5.5-.el7.x86_64
>
>
>
>
> -----Original Message-----
> From: Daniel Gryniewicz [mailto:dang@redhat.com]
> Sent: maandag 30 april 2018 15:15
> To: support(a)lists.nfs-ganesha.org
> Subject: [Nfs-ganesha-support] Re: Nfs-ganesha rgw config for multi
> tenancy rgw users
>
> And test$tester1 is the user ID you created, that is this gives
output:
>
> radosgw-admin user info --uid="test$tester1"
>
> (I'm pounding on this because it really seems that the UID doesn't
> have permissions. If it does, I think there's some really strange
> bug here that's going to be hard to track down.)
>
> Daniel
>
>
> On 04/30/2018 03:25 AM, Marc Roos wrote:
>> Yes same access key and secret access key ( no uid is used )
>>
>> -----Original Message-----
>> From: Daniel Gryniewicz [mailto:dang@redhat.com]
>> Sent: maandag 30 april 2018 14:40
>> To: Marc Roos; support
>> Subject: Re: [Nfs-ganesha-support] Re: Nfs-ganesha rgw config for
>> multi tenancy rgw users
>>
>> I mean does the s3cmd (or cyberduck) use the exact same credentials
>> as
>
>> are specified in the ganesha.conf?
>>
>> Daniel
>>
>> On 04/26/2018 10:09 AM, Marc Roos wrote:
>>>
>>> Yes, I can access the bucket via the s3cmd and cyberduck, or are
>>> you refering to something else?
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: Daniel Gryniewicz [mailto:dang@redhat.com]
>>> Sent: donderdag 26 april 2018 15:38
>>> To: support(a)lists.nfs-ganesha.org
>>> Subject: [Nfs-ganesha-support] Re: Nfs-ganesha rgw config for multi
>>> tenancy rgw users
>>>
>>> Does test$tester1 have permissions to access RGW? Mounting /
>>> should see all the buckets, so it's likely a permission issue.
>>>
>>> Daniel
>>>
>>> On 04/26/2018 03:56 AM, Marc Roos wrote:
>>>>
>>>> I have the same as with / path, I can do the mount, but it is just
>>>> empty
>>>>
>>>> [@mnt]# mount nfs:/rgwtester nfstest/ [@mnt]# ls -l nfstest/
>>>> ls: reading directory nfstest/: Remote I/O error total 0
>>>>
>>>>
>>>>
>>>> Apr 26 14:56:15 c01 nfs-ganesha[3981596]: [main]
>>>> export_commit_common
>>
>>>> :CONFIG :INFO :Export 301 created at pseudo (/rgwtester) with path
>>>> (test3) and tag ((null)) perms (options=03303002 ,
> ,
>>>> , , , , ,
>>>> )
>>>> Apr 26 14:56:15 c01 nfs-ganesha[3981596]: [main]
>>>> export_commit_common
>>
>>>> :CONFIG :INFO :Export 301 created at pseudo (/rgwtester) with path
>>>> (test3) and tag ((null)) perms (options=03303002 ,
> ,
>>>> , , , , ,
>>>> )
>>>> Apr 26 14:56:15 c01 nfs-ganesha[3981596]: [main]
>>>> export_commit_common
>>
>>>> :CONFIG :INFO :Export 301 has 1 defined clients Apr 26 14:56:15
>>>> c01
>>>> nfs-ganesha[3981596]: [main] export_commit_common :CONFIG :INFO
>>>> :Export 301 has 1 defined clients Apr 26 14:56:15 c01
>>>> nfs-ganesha[3981596]: [main] build_default_root :CONFIG :INFO
>>>> :Export
>>
>>>> 0 (/) successfully created Apr 26 14:56:15 c01
nfs-ganesha[3981596]:
>>>> [main] build_default_root :CONFIG :INFO :Export 0 (/)
successfully
>>>> created Apr 26 14:56:16 c01 nfs-ganesha[3981596]:
[main]
>>>> fsal_save_ganesha_credentials :FSAL :INFO :Ganesha uid=0 gid=0
>>>> ngroups=0 Apr 26 14:56:16 c01 nfs-ganesha[3981596]: [main]
>>>> fsal_save_ganesha_credentials :FSAL :INFO :Ganesha uid=0 gid=0
>>>> ngroups=0
>>>>
>>>> -----Original Message-----
>>>> From: Daniel Gryniewicz [mailto:dang@redhat.com]
>>>> Sent: donderdag 26 april 2018 14:47
>>>> To: support(a)lists.nfs-ganesha.org
>>>> Subject: [Nfs-ganesha-support] Re: Nfs-ganesha rgw config for
>>>> multi tenancy rgw users
>>>>
>>>> I think it should be:
>>>>
>>>> Path="test3"
>>>>
>>>> so without the initial '/'. Buckets aren't paths, we're
just
>>>> abusing
>>
>>>> the existing Ganesha config to select them.
>>>>
>>>> Other than that, the client entry should match the equivalent
>>>> entry in
>>>
>>>> your ceph.conf.
>>>>
>>>> Daniel
>>>>
>>>> On 04/26/2018 01:52 AM, Marc Roos wrote:
>>>>>
>>>>>
>>>>> I have problems exporting a bucket that really does exist. I have
>>>>> tried Path = "/test:test3"; Path =
"/test3"; Results in ganesha
>>>>> fails
>>>
>>>>> to start with message ExportId=301 Path=/test:test3
>>>>> FSAL_ERROR=(Invalid object
>>>>> type,0)
>>>>>
>>>>> If I use path=/ I can mount something but that is a empty export,
>>>>> but
>>>
>>>>> cannot put files there.
>>>>>
>>>>> I have this in ganesha (client.rwg3 works with civetweb), should
>>>>> be
>
>>>>> sufficient not?
>>>>>
>>>>> RGW {
>>>>> cluster = "ceph";
>>>>> name = "client.rgw3";
>>>>> ceph_conf = "/etc/ceph/ceph.conf";
>>>>> # for vstart cluster, name = "client.admin"
>>>>> #init_args = "-d --debug-rgw=16"; }
>>>>>
>>>>> EXPORT {
>>>>> Export_ID=301;
>>>>> Path = "/test3";
>>>>> Pseudo = "/rgwtester";
>>>>>
>>>>> FSAL { Name = RGW; User_Id = "test$tester1";
>>>>> Access_Key_Id ="sameass3"; Secret_Access_Key =
"sameass3"; }
>>>>> Disable_ACL = FALSE;
>>>>> CLIENT { Clients = 192.168.x.0/24; } }
>>>>>
>>>>> [@~]$ s3cmd -c .s3cfg.tester1 ls
>>>>> 2018-01-31 21:48 s3://test
>>>>> 2018-02-01 11:44 s3://test2
>>>>> 2018-02-02 17:10 s3://test3
>>>>>
>>>>>
>>>>> [@~]$ s3cmd -c .s3cfg.tester1 ls s3://test3
>>>>> 2018-02-02 17:13 10485760 s3://test3/10MB.txt
>>>>> 2018-02-05 12:57 26784 s3://test3/aB8q5BA_460s.jpg
>>>>>
>>>>> [@ ganesha]# rpm -qa | grep ganesh
>>>>> nfs-ganesha-xfs-2.5.5-.el7.x86_64
>>>>> nfs-ganesha-2.5.5-.el7.x86_64
>>>>> nfs-ganesha-rgw-2.5.5-.el7.x86_64
>>>>> nfs-ganesha-ceph-2.5.5-.el7.x86_64
>>>>> nfs-ganesha-vfs-2.5.5-.el7.x86_64
>>>>> nfs-ganesha-mem-2.5.5-.el7.x86_64
>>>>>
>>>>> On CentOS7
>>>>> _______________________________________________
>>>>> Support mailing list -- support(a)lists.nfs-ganesha.org To
>>>>> unsubscribe
>>
>>>>> send an email to support-leave(a)lists.nfs-ganesha.org
>>>>>
>>>> _______________________________________________
>>>> Support mailing list -- support(a)lists.nfs-ganesha.org To
>>>> unsubscribe
>
>>>> send an email to support-leave(a)lists.nfs-ganesha.org
>>>>
>>>> _______________________________________________
>>>> Support mailing list -- support(a)lists.nfs-ganesha.org To
>>>> unsubscribe
>
>>>> send an email to support-leave(a)lists.nfs-ganesha.org
>>>>
>>> _______________________________________________
>>> Support mailing list -- support(a)lists.nfs-ganesha.org To
>>> unsubscribe send an email to support-leave(a)lists.nfs-ganesha.org
>>>
>>>
>>
>>
>> _______________________________________________
>> Support mailing list -- support(a)lists.nfs-ganesha.org To unsubscribe
>> send an email to support-leave(a)lists.nfs-ganesha.org
>>
> _______________________________________________
> Support mailing list -- support(a)lists.nfs-ganesha.org To unsubscribe
> send an email to support-leave(a)lists.nfs-ganesha.org
>
>