[prev in list] [next in list] [prev in thread] [next in thread] 

List:       nosql
Subject:    Re: Completely unscientific benchmarks :)
From:       "Paul L. Lussier" <plussier () BayNetworks ! COM>
Date:       1999-01-09 14:02:26
[Download RAW message or body]


In a message dated: Sat, 09 Jan 1999 09:22:18 +0100
Carlo Strozzi said:

>On Fri, Jan 08, 1999 at 04:29:11PM -0500, Paul L. Lussier wrote:
> >
> >Don't know if anyone's interested or not:
> >
> >File:
> >	$ ls -l office.lst
> >	-rw-r-----   1 plussier sysadmin    3117 Jan  8 16:13 office.lst
> >
> >OS              SunOS 5.5.1
> >model_name      SUNW,Ultra-5_10
> >kernel_arch     sun4u
> >main_memory     128MB
> >swap            512MB
> >
> >	$ time nosql listtotable < office.lst | nosql sort Data | \
> >	nosql print > /dev/null
> >
> >	real    0m0.55s
> >	user    0m0.23s
> >	sys     0m0.25s
> >	
> >
> >OS              Linux 2.0.36
> >model_name      Intel
> >kernel_arch     i586
> >main_memory     124MB
> >swap            127MB
> >
> >	$ time nosql listtotable < office.lst | nosql sort Data | \
> >	nosql print > /dev/null
> >
> >	0.07 user
> >	0.03 system
> >	0:00.26 elapsed
> >	37%CPU (0avgtext+0avgdata 0maxresident)k
> >	0 inputs + 0 outputs (319major+113minor)pagefaults
> >	0 swaps
> >
>
>Oh, this is interesting :-). Just to make sure it is not something
>related to different implementations of the 'time' shell builtin,
>try with /usr/bin/time on both systems.

Okay, I'll give that a shot.  BTW, Linux system is a P233/32bit CPU,
the Sun is an UltraSparc 270MHz/64bit CPU.  Though I don't think
32 bit vs. 64 bit is an issue, since I think all the processes withing NoSql
are 32 bit, regardless of OS (i.e., awk, perl, etc. are all written
for 32 bit).
-- 

Seeya,
Paul
----
plussier@baynetworks.com
Broadband Technology Division - Bay Networks (now a Nortel Company, Eh? :)

	If you're not having fun, you're not doing it right!

[prev in list] [next in list] [prev in thread] [next in thread]