Thank you Brian Chrisman and Gregory Farnum for help. I understood the concept and able to figure out and fix it by separating journal from data volume.(having a separate partition for journal) On Tue, Mar 8, 2011 at 11:49 PM, Gregory Farnum <gregory.farnum@xxxxxxxxxxxxx> wrote: > On Tuesday, March 8, 2011 at 10:09 AM, Brian Chrisman wrote: > Just making sure I'm interpreting this right. >> At steady state (data_size >> journal_size), total space usage for 2x >> replication should be 2 * data_size + journal_size? >> If you have (data_size < journal_size) and journal is on a file rather >> than on a partition, then btrfs will report 4x data size. > > Steady state should be 2 * data_size + journal_size*num_osds -- assuming all your OSDs have the same size journal, anyway! :) > I'm not sure how the journal grows as a file before it reaches maximum size -- either it starts out as max size (same formula as above), or else it grows as data is written in, yielding for (data_size < journal_size) used_size = data_size * 4. > > Though this obscures the impact of, eg, the metadata -- which will get written to both the MDS journal (stored on OSDs) and to the inodes (in the metadata pool, on OSDs). > -Greg > > > > -- Thanks and Regards, Upendra.M -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
