[ceph-users] jewel - rgw blocked on deep-scrub of bucket index pg

2017-05-05 Thread Sam Wouters
Hi, we have a small cluster running on jewel 10.2.7; NL-SAS disks only, osd data and journal co located on the disks; main purpose rgw secondary zone. Since the upgrade to jewel, whenever a deep scrub starts on one of the rgw index pool pg's, slow requests start piling up and rgw requests are blo

Re: [ceph-users] How does ceph pg repair work in jewel or later versions of ceph?

2017-05-05 Thread David Turner
This was covered in depth in the last 3 months on this ML thread. http://lists.ceph.com/pipermail/ceph-users-ceph.com/2017-February/016373.html On Fri, May 5, 2017, 2:58 AM shadow_lin wrote: > I have read that the pg repair is simply copy the data from the primary > osd to other osds.Is that tr

[ceph-users] Installing pybind manually from source

2017-05-05 Thread Henry Ngo
It appears that the pybind folder from git source needs to be installed manually if building Ceph from source. Where exactly does it need to be copied in Centos? I have tried /usr/local/lib, /usr/local/lib/python2.7/ and /usr/local/lib/python2.7/site-packages with no luck. http://tracker.ceph.com/

Re: [ceph-users] Changing replica size of a running pool

2017-05-05 Thread Alejandro Comisario
Thanks David! Any one ? more thoughts ? On Wed, May 3, 2017 at 3:38 PM, David Turner wrote: > Those are both things that people have done and both work. Neither is > optimal, but both options work fine. The best option is to definitely just > get a third node now as you aren't going to be gett

[ceph-users] RGW: removal of support for fastcgi

2017-05-05 Thread Yehuda Sadeh-Weinraub
RGW has supported since forever. Originally it was the only supported frontend, and nowadays it is the least preferred one. Rgw was first developed over fastcgi + lighttpd, but there were some issues with this setup, so we switched to fastcgi + apache as our main supported configuration. This was

Re: [ceph-users] RGW: removal of support for fastcgi

2017-05-05 Thread Roger Brown
I'm using fastcgi/apache2 instead of civetweb (centos7) because i couldn't get civetweb to work with SSL on port 443 and in a subdomain of my main website. On Fri, May 5, 2017 at 1:51 PM Yehuda Sadeh-Weinraub wrote: > RGW has supported since forever. Originally it was the only supported > front

Re: [ceph-users] RGW: removal of support for fastcgi

2017-05-05 Thread Roger Brown
I'm using fastcgi/apache2 instead of civetweb (centos7) because i couldn't get civetweb to work with SSL on port 443 and in a subdomain of my main website. So I have domain.com, www.domain.com, s3.domain.com (RGW), and *. s3.domain.com for the RGW buckets. As long as you can do the same with civitw

Re: [ceph-users] corrupted rbd filesystems since jewel

2017-05-05 Thread Stefan Priebe - Profihost AG
Hello Json, while doing further testing it happens only with images created with hammer and that got upgraded to jewel AND got enabled exclusive lock. Greets, Stefan Am 04.05.2017 um 14:20 schrieb Jason Dillaman: > Odd. Can you re-run "rbd rm" with "--debug-rbd=20" added to the > command and pos