user-manual: rework IO performance tuning

This commit is contained in:
Thomas Schoebel-Theuer 2019-09-10 10:01:51 +02:00 committed by Thomas Schoebel-Theuer
parent c14b288a4a
commit b88f787c24
1 changed files with 28 additions and 2 deletions

View File

@ -25070,6 +25070,12 @@ re-ordering of writeback
transaction logfile, and even when both are residing at the same RAID set.
\end_layout
\begin_layout Standard
Since RAID-6 is has more expensive write operations due to its CRC computations
and updates, RAID-6 will much more profit from this effect.
Other RAID modes like RAID-10 may show a lower throughput improvement.
\end_layout
\begin_layout Standard
Notice: this effect is not only dependent from total RAM size and from the
maximum size of the MARS temporary memory buffer (tuning parameter
@ -25098,6 +25104,17 @@ dd
is no appropriate benchmarking tool, and has almost nothing to do with
real workloads occuring in datacenters, which typically are neither sequential,
nor do they have a parallelism degree of only 1.
\end_layout
\begin_layout Standard
\noindent
\begin_inset Graphics
filename images/MatieresCorrosives.png
lyxscale 50
scale 17
\end_inset
Please don't try to lead any discussions about this: simply use
\family typewriter
blktrace
@ -25123,11 +25140,20 @@ dd
\end_inset
Any assumptions about workloads are very dangerous: they can deviate from
practice not only by factors, but sometimes even by
Any
\series bold
assumptions about workloads
\series default
are very dangerous: they can deviate from practice not only by factors,
but sometimes even by
\emph on
orders of magnitude
\emph default
.
There no substitute for real measuments of
\series bold
actual workload behaviour
\series default
.
\end_layout