| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 660.1 | I don't think it's different... | PYRO::RON | Ron S. van Zuylen | Wed Aug 09 1995 04:04 | 11 |
| 660.2 | it isn't any different | EEMELI::BACKSTROM | bwk,pjp;SwTools;pg2;lines23-24 | Wed Aug 09 1995 10:46 | 5 |
| 660.3 | | CHEFS::grantt.reo.dec.com::GRANTT | | Wed Aug 09 1995 13:05 | 5 |
| 660.4 | | BULEAN::BANKS | | Wed Aug 09 1995 13:16 | 3 |
| 660.5 | Cluster sizes and effects... | PCBUOA::PDICKERSON | Philip, an Aussie in PC-land | Wed Aug 09 1995 14:16 | 136 |
| 660.6 | More info | TOMMII::RDAVIES | Amateur Expert | Wed Aug 09 1995 14:26 | 34 |
| 660.7 | | CHEFS::grantt.reo.dec.com::GRANTT | | Wed Aug 09 1995 14:34 | 3 |
| 660.8 | | ICS::BEAN | Attila the Hun was a LIBERAL! | Wed Aug 09 1995 18:04 | 6 |
| 660.9 | You don't have to waste the space! | CADSYS::COOK | Alpha - for accurate results | Wed Aug 09 1995 19:33 | 16 |
| 660.10 | "disk compression" without compression is the answer | HYDRA::catherine.crl.dec.com::ALDEN | Ken Alden | Thu Aug 17 1995 17:24 | 6 |
| 660.11 | | BULEAN::BANKS | | Mon Aug 28 1995 19:26 | 6 |
| 660.12 | | LANDO::EIBEN | | Tue Aug 29 1995 19:34 | 9 |
| 660.13 | | MSBCS::BELANGER | Clever quotation goes here ... | Wed Aug 30 1995 11:36 | 5 |
| 660.14 | Opinions Wanted | TAEC::SMITH | Martin Smith, Valbonne. - 828 5128 | Mon Feb 03 1997 05:48 | 7 |
| 660.15 | | BHAJEE::JAERVINEN | Ora, the Old Rural Amateur | Mon Feb 03 1997 07:03 | 9 |
| I have divided my disks in partitions slightly smaller than 512MB (a
compromise between cluster size and having zillions of partitions).
I also tried drivespace, but at least in my case it caused a noticeable
slowdown.
You could of course try to get your hands on Windows 95 OSR2 which
supports FAT32 - 4k clusters up to (I think) 8GB partitions.
|
| 660.16 | OSR2 max partition size ?? | snooty.uvo.dec.com::TRAVELL | John T, UK VMS System Support | Mon Feb 03 1997 11:57 | 9 |
| > You could of course try to get your hands on Windows 95 OSR2 which
> supports FAT32 - 4k clusters up to (I think) 8GB partitions.
Should this not be 16Tb (terabytes!)?
I would expect to see upto 2**32 entries in the FAT, with each entry pointing
to a 4K block. This adds uo to 4294967296 * 4096 = a very very big number :-)
JT:
|
| 660.17 | | BHAJEE::JAERVINEN | Ora, the Old Rural Amateur | Mon Feb 03 1997 12:11 | 5 |
| No, I think I read a few bits are reserved, so over 8GB the cluster
size increases. It's also only supported for partitions > 512MB, i.e.
even if you have OSR2, smaller partitions are formatted as FAT16 (with
larger ones, you can choose either).
|
| 660.18 | opinion offered | DANGER::ARRIGHI | Life is an else-if construct | Mon Feb 03 1997 14:18 | 17 |
| re .14:
I'll tell you what I did on a 3.2GB drive, but it isn't the only way:
drive C: Just under 1024K (16k cluster size)
Lots of stuff likes to go in drive C. Split this
into two if your C drive is on another disk.
drives D, E, F: Just under 512K (8k cluster size)
drive G: The remainder (~< 500K)
drive H: ~100K for the swap file
Tony
|
| 660.19 | | COOKIE::FROEHLIN | Let's RAID the Internet! | Mon Feb 03 1997 22:52 | 7 |
| Swapfile on the last partition? Hope you have enough memory or else
this disk gets a lame arm. Think about, it has to go quite frequently
to C (the first partition on the drive) to H (the last partition).
Especially with the way Windows 95 uses all available free memory for
disk caching you do see swap/paging.
Guenther (always wandering about the random busy LED acitvities)
|
| 660.20 | | DANGER::ARRIGHI | and miles to go before I sleep | Tue Feb 04 1997 15:15 | 7 |
| Hmmmm..... Well I've got 48MB, but you have a point. So, what would
be nice, would be a way to change the mapping of the partition letters
to the disk surface.
(And, by the way, change all my K's in .18 to M's.)
Tony
|
| 660.21 | Just a Quick Question. | TAEC::SMITH | Martin Smith, Valbonne. - 828 5128 | Wed Feb 05 1997 10:07 | 8 |
| 660.22 | | TARKIN::LIN | Bill Lin | Wed Feb 05 1997 10:41 | 18 |
| re: .21 by TAEC::SMITH
>> What if C: was a separate disk, and the swapfile was on the last
>> partition of a second (bigger (3Gb?)) disk. Surely that can be a
>> benefit?
Hi Martin,
My gut feel says yes. The chances are also good that your 3GB disk is
faster than your original/boot disk. And if you're running SCSI, there
would be the additional benefit that your spindles can run separately
because of command queuing (sp?).
Does (E)IDE support command queuing yet? Anyone know for certain?
Cheers,
/Bill
|
| 660.23 | Small drive for browsers | SUBSYS::CHESTER | | Fri Feb 07 1997 16:06 | 13 |
| Another way of spliting a 2g drive.
C: 502 MB Win95, office and any sw that must be installed on C:
D: 502 MB Any sw that allows installs to your choice of drives
E: 50MB Netscape, AOL. The 1kb cluster works well with the
thousands of 1k cache files these programs generate
F: 950 MB Mostly data files and other SW.
I leave E: out of the backups. Arcada backup was have "heart failure"
trying to catalog all the cache files.
Kc
|
| 660.24 | | BHAJEE::JAERVINEN | Ora, the Old Rural Amateur | Fri Feb 07 1997 16:11 | 7 |
| re .23:
>C: 502 MB Win95, office and any sw that must be installed on C:
Why would you want to install Office on C:? Mine's on W:
|
| 660.25 | | BULEAN::BANKS | Orthogonality is your friend | Wed Feb 12 1997 12:12 | 12 |
| Re: .23:
I was experiencing Arcada backup heart failures on my netscape cache as
well. One day I just got frustrated and left the room.
The amazing thing was that it actually completed. It just wanted to spend
10-15 minutes (at the time, on a P-120!) recataloging the cache.
Since then, I've learned to just delete the catalog files before doing a
backup. Saves space on the catalog drive (since it's a huge file, that
just gets huger), and it makes the backups run marginally faster, even when
it isn't losing its mind to the netscape cache.
|