Recently we tried to figure out if there was a risk in activating the commit cache. It seems impossible to get a description as to how this cache works. Talking to my NSC I got the impression that it is fairly safe to use it but I'm still not sure. What if the server crashes while there is still some data in the cache? What information is stored in commit cache? And in what order is this information written to disk?
Anyway, performance is such an issue we decided to activate commit cache.
Nice thing is we have a tool (client side report) which merely uses the disk performance tool to write the read/write times to a logfile.
And now, the curious part:
- Before using commit cache:
- read avg: 7 msec (with peaks up to 30 msec)
- write avg: 9 msec (with peaks up to 100 msec)
- After enabling commit cache:
- read avg: 9 msec (with peaks up to 30 msec)
- write avg: 19 msec (with peaks up to 100 msec)
Lukily, the end-user is perceiving a slight performance improvement. But I'm puzzled, are there tools to get a measurable indicator for the performance?
While I'm at it; anyone suggestions for some good reading on Navision performance tuning?
So far I've been looking into "Installation & System management" (w1w1ism.pdf) and of course this forum.
Comments
MVP - Dynamics NAV
My BLOG
NAVERTICA a.s.
That commit cache parameter is a great difference. You do not need a stopwatch to notice that.
www.dasautomatisering.nl