Re: corruption with vfp 8 - continued
- From: Jay B <jayB@xxxxxxxxxxx>
- Date: Sat, 29 Apr 2006 13:50:51 -0400
thanks thomas.
wow, i still cant believe that with all the memory and resources in these modern computers, this is still an issue here.
Thomas Ganss wrote:
Hi Jay,.
there still are some memory issues in vfp8 and vfp9.
I encountered a few 1150's in a reproducable manner
in certain levels of sys(3050) settings.
The general rule seems to be to set sys(3050) to something below
192 MB even on very well equipped machines
(ranging from 512 MB to 2 GB,
Swap file size = Max(1.5GB, 1.5*PhysicalMem) )
Verified in vfp8Sp1, vfp9Sp0, vfp9Sp1.
The funny thing is that there seems to be no fail/works-limit:
for instance the 2 Gig machines would
work at 192
fail at 250MB,
work at 390MB,
fail at 500MB
work at default setting (nearly 700MB)
and I've not found a pattern across every mem setting.
Since the "my" problematic app has it's roots back in 2.x
it is mostly xBase code - seek, scan, locate - which is
often quicker when running in fairly limited memory space,
I was able to minimize sys(3050) in most of the program.
I'ld reccommend setting memory to someting like
min(PhysicalRAM*0.2, 112) MB via setup method.
If there are areas the app is working poorly because of starved mem,
(SQL Joins, reindexes and so on) bump up before and lower down afterwards.
And make sure that the area each user's tempfiles are
created in is large enough.
One other thing: I had 43's crop up on one machine where
one of the disks ran at high temperature (too many disks
and a hot running PIV). Not using that disk helped a lot <g>
my 0.02 EUR
thomas
- References:
- corruption with vfp 8 - continued
- From: Jay B
- Re: corruption with vfp 8 - continued
- From: swdev2
- Re: corruption with vfp 8 - continued
- From: Jay B
- Re: corruption with vfp 8 - continued
- From: Thomas Ganss
- corruption with vfp 8 - continued
- Prev by Date: Re: corruption with vfp 8 - continued
- Next by Date: Re: DBF to Excel File, 16383 is the limit ?
- Previous by thread: Re: corruption with vfp 8 - continued
- Next by thread: Re: corruption with vfp 8 - continued
- Index(es):
Relevant Pages
|