<div dir="ltr"><div>All,<br></div><div style><br></div><div style>I have been observing some odd behavior regarding write throughput to an XFS partition (the baseline kernel version is 2.6.32.27). I see consistently high write throughput (close to the performance of the raw block device) to the filesystem immediately after a mkfs, but after a few test cycles, there is sporadic poor performance.</div>
<div style><br></div><div style>The test mechanism is like so:</div><div style><br></div><div style>[mkfs.xfs <blockdev>] (no flags/options, xfsprogs ver 3.1.1-0.1.36)</div><div style>...</div><div style>1. remove a previous test cycle's directory </div>
<div style>2. create a new directory</div><div style>3. open/write/close a small file (4kb) in this directory</div><div style>4. open/read/close this same small file (by the local NFS server)</div><div style>5. open[O_DIRECT]/write/write/write/.../close a large file (anywhere from ~100MB to 200GB)</div>
<div style><br></div><div style>Step #5 contains the high-throughput metrics which becomes an order of magnitude worse several test cycles after a mkfs. Omitting steps 1-3 does not show the poor performance behavior.</div>
<div style><br></div><div style>Can anyone provide any suggestions as to an explanation for the behavior or a way to mitigate it? Running xfs_fsr didn't seem to improve the results.</div><div style><br></div><div style>
I'm happy to share benchmarks, specific results data, or describe the hardware being used for the measurements if it's helpful.</div><div><div><br></div>-- <br>-Brian
</div></div>