On Thu, 13 Oct 2016, Somnath Roy wrote:
> Sage,
> I found the following counters are tracking (from _flush_range()) the number of WAL and SST writes happening in rocksdb. Based on the today's discussion on characterizing rocksdb write pattern, isn't these two counters are helpful ?
>
> l_bluefs_bytes_written_wal
> l_bluefs_bytes_written_sst
Maybe, but I *think* want we want to know is how much wal data is ending
up in the level0 (or later) sst files. I don't think there is a way to do
that except to go and look at the level0 files and see how many key/value
data has the wal prefix ("L\0"). I don't think it's even very easy to
tell from bluefs files which sst files are l0.. probably need to parse the
manifest to see that. This should be possible by exporting the bluefs
data, though, which should be possible with the ceph-bluefs-tool.
sage
--
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
