On 2013-04-25 09:39, James Harper wrote:
I'm doing some testing and wanted to see the effect of increasing journal speed, and the fastest way to do this seemed to be to put it on a ramdisk where latency should drop to near zero and I can see what other inefficiencies exist. I created a tmpfs of sufficient size, copied journal on to that, and started the OSD. It core dumped pretty quickly which I assume is related to direct io not being supported. I turned off journal dio and the osd started okay but never recovered. How should I do this? Is there another type of ramdisk I should use or other journal options I should specify? (and yes I understand that losing ramdisk will lose the osd - I'm just doing testing)
I'm currently doing the same, for evaluation purposes. I tried copying the journal, but never got it working properly. In the end, I just created some new OSDs with journals on tmpfs and journal dio = false.
-- Carl-Johan Schenström Driftansvarig Språkbanken & Svensk nationell datatjänst, Göteborgs universitet E-post: carl-johan.schenstrom@xxxxx | Mobil: +46 709 116769 _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
