<font size=2 face="sans-serif">Its also important to look into the actual
space "wasted" by the "subblock mismatch".</font><br><font size=2 face="sans-serif">For example, a snip from a filehist
output I've found somewhere:</font><br><br><font size=2 face="sans-serif"><i>File%ile represents the cummulative
percentage of files.</i></font><br><font size=2 face="sans-serif"><i>Space%ile represents the cummulative
percentage of total space used.</i></font><br><font size=2 face="sans-serif"><i>AvlSpc%ile represents the cummulative
percentage used of total available space.</i></font><br><br><font size=2 face="sans-serif"><i>Histogram of files <= one 2M block
in size</i></font><br><font size=2 face="sans-serif"><i>Subblocks Count File%ile
Space%ile AvlSpc%ile</i></font><br><font size=2 face="sans-serif"><i>--------- -------- ---------- ----------
----------</i></font><br><font size=2 face="sans-serif"><i> 0 1297314
2.65% 0.00%
0.00%</i></font><br><font size=2 face="sans-serif"><i> 1 34014892
72.11% 0.74% 0.59%</i></font><br><font size=2 face="sans-serif"><i> 2 2217365
76.64% 0.84% 0.67%</i></font><br><font size=2 face="sans-serif"><i> 3 1967998
80.66% 0.96% 0.77%</i></font><br><font size=2 face="sans-serif"><i> 4
798170 82.29% 1.03%
0.83%</i></font><br><font size=2 face="sans-serif"><i> 5 1518258
85.39% 1.20% 0.96%</i></font><br><font size=2 face="sans-serif"><i> 6
581539 86.58% 1.27%
1.02%</i></font><br><font size=2 face="sans-serif"><i> 7
659969 87.93% 1.37%
1.10%</i></font><br><font size=2 face="sans-serif"><i> 8 1178798
90.33% 1.58% 1.27%</i></font><br><font size=2 face="sans-serif"><i> 9
189220 90.72% 1.62%
1.30%</i></font><br><font size=2 face="sans-serif"><i> 10
130197 90.98% 1.64%
1.32%</i></font><br><br><br><font size=2 face="sans-serif">So, 72% of the files are smaller then
1 subblock ( 2M in the above case BTW). If, for example, we'll double it
- we will "waste" ~76% of the files, and if we'll push it to
16M it will be ~90% of the files...</font><br><font size=2 face="sans-serif">But, we really care about capacity,
right? So, going into the 16M extreme, we'll "waste" 1.58% of
the capacity ( worst case of course).</font><br><br><font size=2 face="sans-serif">So, if it will give you ( highly depends
on the workload of course) 4X the performance ( just for the sake of discussion)
- will it be OK to pay the 1.5% "premium" ?</font><br><br><br><font size=2 face="sans-serif"><br>Regards,<br><br>Tomer Perry<br>Scalable I/O Development (Spectrum Scale)<br>email: tomp@il.ibm.com<br>1 Azrieli Center, Tel Aviv 67021, Israel<br>Global Tel: +1 720 3422758<br>Israel Tel: +972 3 9188625<br>Mobile: +972 52 2554625<br></font><br><br><br><br><font size=1 color=#5f5f5f face="sans-serif">From:
</font><font size=1 face="sans-serif">"Marc A Kaplan"
<makaplan@us.ibm.com></font><br><font size=1 color=#5f5f5f face="sans-serif">To:
</font><font size=1 face="sans-serif">gpfsug main discussion
list <gpfsug-discuss@spectrumscale.org></font><br><font size=1 color=#5f5f5f face="sans-serif">Date:
</font><font size=1 face="sans-serif">10/04/2019 20:57</font><br><font size=1 color=#5f5f5f face="sans-serif">Subject:
</font><font size=1 face="sans-serif">Re: [gpfsug-discuss]
Follow-up: ESS File systems</font><br><font size=1 color=#5f5f5f face="sans-serif">Sent by:
</font><font size=1 face="sans-serif">gpfsug-discuss-bounces@spectrumscale.org</font><br><hr noshade><br><br><br><font size=2>If you're into pondering some more tweaks:</font><font size=3><br></font><font size=2><br>-i InodeSize is tunable</font><font size=3><br></font><font size=2><br>system pool : --metadata-block-size is tunable separately from -B
blocksize</font><font size=3><br></font><font size=2><br>On ESS you might want to use different block size and error correcting
codes for (v)disks that hold system pool.<br>Generally I think you'd want to set up system pool for best performance
for relatively short reads and updates.<br></font><font size=3><br></font><tt><font size=2>_______________________________________________<br>gpfsug-discuss mailing list<br>gpfsug-discuss at spectrumscale.org<br></font></tt><a href="http://gpfsug.org/mailman/listinfo/gpfsug-discuss"><tt><font size=2>http://gpfsug.org/mailman/listinfo/gpfsug-discuss</font></tt></a><tt><font size=2><br></font></tt><br><br><BR>