Hi Mark,
I think you are the right man for these questions :) I am really don't
understand how osd_client_message_size_cap , objecter_infilght_op_bytes/ops,
ms_dispatch_throttle_bytes works? And how they affect performance.
Especially ,the objecter_inflight_op_bytes seems be used in the osdc,
well, I don't know what osdc means, but just assuming it's client related.I
cannot understand how tuning in ceph side be passed to client side?
The situation we met is, our disks(both journal and data disks) are far
from busy enough, but when I enable osd debugging, from the log, I could see
the journal queue is usually empty.this indicate that the client cannot push
enough workload to ceph.we are using 48 VMs with 48 RBDs,and ceph totally have
80 spindles, I suppose the workload is heavy enough.
Xiaoxi
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
