home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!rutgers!cbmvax!mks
- From: mks@cbmvax.commodore.com (Michael Sinz)
- Newsgroups: comp.sys.amiga.hardware
- Subject: Re: 4 Gigs....and counting..
- Keywords: SCSI Amigados
- Message-ID: <34263@cbmvax.commodore.com>
- Date: 14 Aug 92 15:18:17 GMT
- References: <1992Aug7.140551.25211@walter.cray.com> <1992Aug9.034453.4560@msuinfo.cl.msu.edu> <32648@auc.UUCP>
- Reply-To: mks@cbmvax.commodore.com (Michael Sinz)
- Organization: Commodore, West Chester, PA
- Lines: 70
-
- rar@auc.UUCP (Rodney Ricks) writes:
- >In article <1992Aug9.034453.4560@msuinfo.cl.msu.edu> jap@crs.cl.msu.edu (Joe Porkka) writes:
- >>Due to the standard AmigaDOS device interface any particular
- >>(logical) device is limited to 4gig. However, you may have many 4gig devices
- >>attached. Theoretically you could have lets see...
- >>In a A3000 which has 4 slots
- >> 5 SCSI controllers, each can get at 7 unit of 8 drives each -
- >> giving you 5*7*8 or 280 drives. Say each drive was 4GIG.
- >> That leaves you with 1120 GIG of disk space.
- >> (You better have one *BIG* powersupply)
- >
- >First of all, is the 4 gig limit on the drives in this case, or the partitions?
- >Aren't the partitions the logical devices? If so, the amount of disk space
- >is only limited by the number of devices that you can have (times the maximum
- >partition size), and I think that is theoretically unlimited in the Amiga OS
- >(if it is limited, I bet the limit occurs at the DOS level.).
-
- The limit is that the device structures/names must fit in memory.
- Other than this, there is no real limit. After all, it is all linked lists.
-
- >Actually, I think the number of logical devices is limited by the naming
- >convention (three characters for the actual device name).
-
- Device names can be up to 30 characters. Mine show up as: (from Info)
-
- Mounted disks:
- Unit Size Used Free Full Errs Status Name
- SOFTSYS: 32M 12875 52715 20% 0 Read/Write SoftSys
- WB: 19M 30024 10224 75% 0 Read/Write WB
- TYBALT: 11M 11326 11842 49% 0 Read/Write Tybalt
- TEST: 681K 1272 91 93% 0 Read/Write Test
- 040: 40M 67514 14419 82% 0 Read/Write 040
- JUNIOR: 29M 18462 41008 31% 0 Read/Write Junior
- SOFT: 1401M 1682481 1188373 59% 0 Read/Write Main
- RAM: 1K 1 0 101% 0 Read/Write Ram Disk
- WB_2.x: 19M 18784 22160 46% 0 Read/Write System
- DF0: No disk present
- DF1: No disk present
- T: 20M 2331 38667 6% 0 Read/Write T
- Mail: 29M 9762 51654 16% 0 Read/Write Mail
- Big: 499M 95434 928498 9% 0 Read/Write Big
- BackUp: 583M 442553 753251 37% 0 Read/Write BackUp
- Trash: 299M 50853 563489 8% 0 Read/Write Trash
- Other: 99M 85835 118833 42% 0 Read/Write P
- Work: 169M 28776 318142 8% 0 Read/Write Work
-
- >Second, isn't the partition size limit under AmigaDOS 2gig? I think I remember
- >that it had a 2gig limit, and that signed arithmetic was used.
-
- Well, we have not yet been able to test the system if it works with larger
- than 2gig. I do know that it will be a problem to have a file over 2gig due
- to the fact that Seek() (and fseek()) take a signed value for the byte offset.
-
- The size of the device (not partition) is limited to 4gig (32-bit) since
- the Amiga .device I/O model takes byte offsets as a 32-bit number. A device
- driver could take a 8gig drive and make it look like two different units of
- 4gig each.
-
- However, I claim that the number of people that will put a drive that is
- over 2gig in size on their $500 Amiga coputer system will be rather low
- for a while yet. By the time this really becomes a problem, the fix will
- have been done long ago.
-
- /----------------------------------------------------------------------\
- | /// Michael Sinz - Senior Amiga Systems Engineer |
- | /// Operating System Development Group |
- | /// BIX: msinz UUNET: mks@cbmvax.commodore.com |
- |\\\/// |
- | \XX/ HELP! I am starting to like it here... |
- \----------------------------------------------------------------------/
-