|
2021-06-11, 00:34
|
#1 |
May 2011
Orange Park, FL
3×5×59 Posts |
Does the PRP proof file cause disk fragmentation?
Does the preallocation of the large PRP proof file at the start of a PRP test contribute to disk fragmentation in Windows? I have watched as this file is created at the start of a run and it grows and grows until it reaches its full required size.
Is prime95 allocating the space all at once, or is it writing a portion at a time to reach the required size? I ask, since I have stopped prime95 a short ways into a new PRP test and run a Diskeeper optimization and there are a large number of free spaces consolidated. Is it recommended to optimize each time a new PRP test created a new proof file? |
|
|
|
2021-06-11, 00:39
|
#2 |
"Oliver"
Sep 2017
Porta Westfalica, DE
72×11 Posts |
Creating and writing a file not on a low level always may cause fragmentation. On Linux at least, one (e.g. mprime) could use fallocate. I do not now which method is used internally.
Fragmentation should not cause problems until the proof will be generated. Even then I am not sure. How many fragments do you count? A few (i.e. < 100) should not cause a major slowdown. Optimizing your system only because of Prime95/mprime is surely not recommended. If that would be the case, George likely would have to do some bug fixing... Last fiddled with by kruoli on 2021-06-11 at 00:40 Reason: Grammar. |
|
|
|
2021-06-11, 00:47
|
#3 | |
|
May 2011
Orange Park, FL
3·5·59 Posts |
Quote:
|
|
|
|
|
|
2021-06-11, 00:54
|
#4 |
|
"Oliver"
Sep 2017
Porta Westfalica, DE
72·11 Posts |
Since the allocation should be done before the first residue is saved, this really might be a problem. But please check this back again.
|
|
|
|
|
2021-06-11, 00:55
|
#5 |
|
Undefined
"The unspeakable one"
Jun 2006
My evil lair
183416 Posts |
IMO disk fragmentation is an overblown non-issue for most uses. Don't worry about it. Companies that make the defragmentation tools will of course say how awful it is and that you definitely need their stuff. But if it really was such a huge problem as they claim then the FS drivers would have it inbuilt.
In cases of extreme fragmentation things might go a bit slower, that is all. If you are really worried about it, once in a few months or so of running a defragmenter should be fine. Last fiddled with by retina on 2021-06-11 at 00:56 |
|
|
|
|
2021-06-11, 01:00
|
#6 | |
|
"Oliver"
Sep 2017
Porta Westfalica, DE
72×11 Posts |
Quote:
At my work, my boss still likes to order HDDs as a boot drive. A lot of loading processes behave like having fragmentation (high counts of small files). I despise him for this. |
|
|
|
|
|
2021-06-11, 01:08
|
#7 |
|
6809 > 6502
"""""""""""""""""""
Aug 2003
101×103 Posts
263A16 Posts |
If I recall correctly, it reserves the whole size and then files in as needed. There was discussion during the early stages about writing the files to do the certs and file allocation and SSD. You might try searching some of the threads about that.
|
|
|
|
|
2021-06-11, 01:33
|
#8 | |
|
May 2011
Orange Park, FL
3×5×59 Posts |
Quote:
|
|
|
|
|
|
2021-06-11, 01:35
|
#9 | |
|
May 2011
Orange Park, FL
3×5×59 Posts |
Quote:
|
|
|
|
|
|
2021-06-11, 02:03
|
#10 |
"Mike"
Aug 2002
202016 Posts |
|
|
|
|
 |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Unexpected error during .proof file upload | Raydex | Software | 4 | 2021-03-08 16:00 |
| How do I get the ~133 mb proof file delivered to GIMPS in the manual environment? | tsc | Information & Answers | 10 | 2020-11-23 23:34 |
| [Help] Cannot open PRP proof interim residues file: n/p108897011.residues | wagner85 | Information & Answers | 5 | 2020-11-16 03:02 |
| memory fragmentation and GMP | Chris Card | Programming | 0 | 2006-12-18 16:09 |
| disk writes and results file outputs | nostream | Software | 4 | 2005-07-16 10:12 |
