Messages in this thread | | | Date | Sun, 21 Mar 1999 01:09:38 +0100 (CET) | From | Jelle Foks <> | Subject | Re: 2.2.3, VIA, IDE, performace -> use bonnie |
| |
On Sat, 20 Mar 1999, Meelis Roos wrote:
> jpcd> Various tests describe this thing as fast (~6.5 MB/sec), and it is the only > jpcd> IDE device I have in my machine, but hdparm only tells me 2.5MB/sec (my > jpcd> DCAS-34330W does 5.8MB/sec according to hdparm, and ~4MB/sec according to > jpcd> magazines). > > Only a little realted (but still): Intel Celeron 366 with some Intel > chipset. Originally had 4G IBM DTTA disk. hdparm showed 11.85 MB/s. > Then the disk was upgraded to a 8G IBM DTTA. Now hdparm reports 5.9..6.2 > M/s on the first run. Slower than it showed for the 4G disk. > > Now the interesting part: running hdparm -t /dev/hda again (and again > and again) shows constantly 84.21 MB/s. This is the part I can't > understand - even if the disk is capable of caching the data, EIDE would > not let more than 33 MB/s through.
RAM lets through more than 33MB/s...
Why don't you use bonnie for your benchmarks? (writes&reads to a file of up to 2GB with different strategies (per char (large cpu overhead), per block (small cpu overhead)), and it does seeks as a benchmark.
Though probably even bonnie can be improved, it's much better than hdparm for quick tests.
Cya, Jelle.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |