Title: BlueSCSI v2 Time Trials
Post by: cooltr6 on February 23, 2024, 02:16:26 AM
Post by: cooltr6 on February 23, 2024, 02:16:26 AM
As our spinning hard drive's continue to age and fail, we look for alternatives to keep our beloved NeXT hardware running happily. There are many worthy devices out there - SCSI2SD/ZuluSCSI and BlueSCSI to name a couple.
This work log brings to you some 'in the wild' testing of BlueSCSI [internal and external versions], SCSI2SD and even a still happily functioning 1GB DEC SCSI hard drive.
The first test is with a NeXTstation Mono Turbo, which has 16MB RAM installed.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
This computer can connect to the Internet and internal network and is set to mount an NFS share, which is on a Raspberry Pi [see NFS for Everyone post - https://www.nextcomputers.org/forums/index.php?topic=5544.0]. This adds a second or so during the startup process...so something to keep in mind.
If you're thinking using BlueSCSI, take a look at the post "BlueSCSI Settings" posted in the Black Hardware section of the forum - https://www.nextcomputers.org/forums/index.php?topic=5498 for some extra info.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 59.5 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 17.5 seconds
Launching
External BlueSCSI v2
From bonk at POST to Login Window: 61.0 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 42.1 seconds
Launching
Note: I don't have a whole lot of applications loaded on this card [yet], so took a few of the big names to test with.
Analysis
Bonk at POST to Login Window: nearly the same times...but certainly hands-down better than a spinning platter drive.
Login Window to Workspace loaded [mounting first CD]: This is where things got interesting. The internal version, the CD mounted as Workspace loaded. With the external version, there was close to 20 seconds difference before the CD mounted on the Shelf. Can't explain that one, but it's something to be aware of.
Application launches were almost dead on the same, except one...Concurrence and Diagram loaded faster with the external version. Other than those two, nothing exactly stood out.
Subsequent Testing
NeXT Cube 25MHz 68040 with 64MB RAM, Dimension board and an internal CD-ROM drive, using both the internal and external versions of BlueSCSI and the same SD card used in the Mono Turbo Slab.
The Cube has a SCSI2SD card installed, with OPENSTEP 4.2, the one used to build the BlueSCSI SD card.
For comparison purposes, I'll take the same time trials booting up the Cube with the two BlueSCSI devices. I'll also take times from booting with the SCSI2SD solution both in the Cube and the Mono Slab. Of course, the SCSI2SD solution doesn't have an external version — but that's OK. Just looking to compare two very worthy SCSI drive replacement options that are out there.
With the other Cube I have, which has less memory, a 1GB DEC SCSI hard drive [with a full install of OPENSTEP 4.2] and no Dimension board installed - run the same tests with BlueSCSI and SCSI2HD. I envision it should be the same as the loaded up Cube, but hey...may as well check. It certainly can't hurt.
While I have the hard drive out of that Cube, I'll do some other testing with that in the loaded Cube, the not-so-loaded Cube and the Mono Slab and get some times...so we can see the 'spinny-drive' performance numbers for comparison purposes.
More to come as the other tests are performed and the details tabulated. Stay tuned!
This work log brings to you some 'in the wild' testing of BlueSCSI [internal and external versions], SCSI2SD and even a still happily functioning 1GB DEC SCSI hard drive.
The first test is with a NeXTstation Mono Turbo, which has 16MB RAM installed.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
This computer can connect to the Internet and internal network and is set to mount an NFS share, which is on a Raspberry Pi [see NFS for Everyone post - https://www.nextcomputers.org/forums/index.php?topic=5544.0]. This adds a second or so during the startup process...so something to keep in mind.
If you're thinking using BlueSCSI, take a look at the post "BlueSCSI Settings" posted in the Black Hardware section of the forum - https://www.nextcomputers.org/forums/index.php?topic=5498 for some extra info.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 59.5 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 17.5 seconds
Launching
WordPerfect 1.0.1: 8.0 seconds
FrameMaker 3.2: 7.6 seconds
Concurrence: 2.77: 10.0 seconds
Diagram 2.5: 14.2 seconds
OpenWrite 2.1-8: 8.3 seconds
ParaSheet 1.7: 16.6 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 61.0 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 42.1 seconds
Launching
WordPerfect 1.0.1: 8.4 seconds
FrameMaker 3.2: 8.1 seconds
Concurrence: 2.77: 7.9 seconds
Diagram 2.5: 13.8 seconds
OpenWrite 2.1-8: 8.9 seconds
ParaSheet 1.7: 16.8 seconds
Note: I don't have a whole lot of applications loaded on this card [yet], so took a few of the big names to test with.
Analysis
Bonk at POST to Login Window: nearly the same times...but certainly hands-down better than a spinning platter drive.
Login Window to Workspace loaded [mounting first CD]: This is where things got interesting. The internal version, the CD mounted as Workspace loaded. With the external version, there was close to 20 seconds difference before the CD mounted on the Shelf. Can't explain that one, but it's something to be aware of.
Application launches were almost dead on the same, except one...Concurrence and Diagram loaded faster with the external version. Other than those two, nothing exactly stood out.
Subsequent Testing
NeXT Cube 25MHz 68040 with 64MB RAM, Dimension board and an internal CD-ROM drive, using both the internal and external versions of BlueSCSI and the same SD card used in the Mono Turbo Slab.
The Cube has a SCSI2SD card installed, with OPENSTEP 4.2, the one used to build the BlueSCSI SD card.
For comparison purposes, I'll take the same time trials booting up the Cube with the two BlueSCSI devices. I'll also take times from booting with the SCSI2SD solution both in the Cube and the Mono Slab. Of course, the SCSI2SD solution doesn't have an external version — but that's OK. Just looking to compare two very worthy SCSI drive replacement options that are out there.
With the other Cube I have, which has less memory, a 1GB DEC SCSI hard drive [with a full install of OPENSTEP 4.2] and no Dimension board installed - run the same tests with BlueSCSI and SCSI2HD. I envision it should be the same as the loaded up Cube, but hey...may as well check. It certainly can't hurt.
While I have the hard drive out of that Cube, I'll do some other testing with that in the loaded Cube, the not-so-loaded Cube and the Mono Slab and get some times...so we can see the 'spinny-drive' performance numbers for comparison purposes.
More to come as the other tests are performed and the details tabulated. Stay tuned!
Title: Re: BlueSCSI v2 Time Trials
Post by: barcher174 on February 24, 2024, 10:59:46 PM
Post by: barcher174 on February 24, 2024, 10:59:46 PM
Very useful comparisons! Looking forward to more.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on February 25, 2024, 07:00:13 PM
Post by: cooltr6 on February 25, 2024, 07:00:13 PM
BlueSCSI Time Trials - The 'Not-so-Loaded Cube'
25MHz 68040
16MB RAM/100ns
Normally has a 1GB DEC SCSI Hard Drive inside; disconnected for this testing.
I will load the same applications onto this drive, so I can get the 'spinny drive' numbers
No Dimension board or other upgrades.
...really a plain Cube overall.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 80.5 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 23.5 seconds
Launching
External BlueSCSI v2
From bonk at POST to Login Window: 80.1 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 48.2 seconds
Launching
Analysis
Much the same as the first set of testing with the Mono Turbo Slab, nearly identical times from bonk at POST to the login window. Same phenomenon of the slowness mounting the first CD in Workspace showed up. I have a feeling this will be the same for the remainder of the tests, but time [no pun intended] will tell.
Interesting comparisons already. In the Mono Turbo Slab testing, Concurrence and Diagram were the two that stood out, but with the 'not-so-loaded' Cube, the times were nearly identical, across Internal and External versions of BlueSCSi.
Nothing really jumped out at me with this set of testing. Let's keep in mind that at the moment, we're comparing apples-to-oranges. 25MHz as opposed to 33MHz. However, I also have a non-Turbo Mono Slab. It too will be used in this range of tests.
Much more to come!
25MHz 68040
16MB RAM/100ns
Normally has a 1GB DEC SCSI Hard Drive inside; disconnected for this testing.
I will load the same applications onto this drive, so I can get the 'spinny drive' numbers
No Dimension board or other upgrades.
...really a plain Cube overall.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 80.5 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 23.5 seconds
Launching
WordPerfect 1.0.1: 11.5 seconds
FrameMaker 3.2: 11.2 seconds
Concurrence: 2.77: 12.8 seconds
Diagram 2.5: 16.1 seconds
OpenWrite 2.1-8: 10.9 seconds
ParaSheet 1.7: 18.9 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 80.1 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 48.2 seconds
Launching
WordPerfect 1.0.1: 11.2 seconds
FrameMaker 3.2: 11.6 seconds
Concurrence: 2.77: 13.2 seconds
Diagram 2.5: 15.1 seconds
OpenWrite 2.1-8: 11.4 seconds
ParaSheet 1.7: 20.1 seconds
Analysis
Much the same as the first set of testing with the Mono Turbo Slab, nearly identical times from bonk at POST to the login window. Same phenomenon of the slowness mounting the first CD in Workspace showed up. I have a feeling this will be the same for the remainder of the tests, but time [no pun intended] will tell.
Interesting comparisons already. In the Mono Turbo Slab testing, Concurrence and Diagram were the two that stood out, but with the 'not-so-loaded' Cube, the times were nearly identical, across Internal and External versions of BlueSCSi.
Nothing really jumped out at me with this set of testing. Let's keep in mind that at the moment, we're comparing apples-to-oranges. 25MHz as opposed to 33MHz. However, I also have a non-Turbo Mono Slab. It too will be used in this range of tests.
Much more to come!
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on February 26, 2024, 01:41:13 AM
Post by: cooltr6 on February 26, 2024, 01:41:13 AM
BlueSCSI Time Trials - NeXTstation Mono Turbo and a Standard Hard Drive
33MHz 68040
16MB RAM/70ns
No other upgrades.
Verbose POST
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 15.5 seconds
From Drive Ready to Login Window: 58.3 seconds
Overall from bonk at POST to Login Window: 73.7 seconds
From Login Window to Workspace loaded: 20.9 seconds
Launching
Return to Login Window from Workspace: 5.9 seconds
Analysis
When comparing startup, the standard hard drive does tend to add some extra time to the process. An extra 15± seconds were added, waiting for the hard drive to actually come ready and begin the actual startup process.
From Login Window to Workspace is nearly the same as the Internal BlueSCSI setup and half of External BlueSCSI.
Launching applications are nearly double the time of Internal and External BlueSCSI. This is not unexpected, as we're dealing with a mechanical drive in this battery of tests.
I added another quick check of times, logging out of Workspace. For the platter drive, 5.9 seconds was not unacceptable. I'll follow up with the other published tests with this same time check.
33MHz 68040
16MB RAM/70ns
No other upgrades.
Verbose POST
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 15.5 seconds
From Drive Ready to Login Window: 58.3 seconds
Overall from bonk at POST to Login Window: 73.7 seconds
From Login Window to Workspace loaded: 20.9 seconds
Launching
WordPerfect 1.0.1: 17.8 seconds
FrameMaker 3.2: 11.2 seconds
Concurrence: 2.77: 13.6 seconds
Diagram 2.5: 14.6 seconds
OpenWrite 2.1-8: 10.5 seconds
ParaSheet 1.7: 17.3 seconds
Return to Login Window from Workspace: 5.9 seconds
Analysis
When comparing startup, the standard hard drive does tend to add some extra time to the process. An extra 15± seconds were added, waiting for the hard drive to actually come ready and begin the actual startup process.
From Login Window to Workspace is nearly the same as the Internal BlueSCSI setup and half of External BlueSCSI.
Launching applications are nearly double the time of Internal and External BlueSCSI. This is not unexpected, as we're dealing with a mechanical drive in this battery of tests.
I added another quick check of times, logging out of Workspace. For the platter drive, 5.9 seconds was not unacceptable. I'll follow up with the other published tests with this same time check.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on February 26, 2024, 02:54:32 AM
Post by: cooltr6 on February 26, 2024, 02:54:32 AM
BlueSCSI Time Trials - The 'Not-so-Loaded Cube' and a Standard Hard Drive
25MHz 68040
16MB RAM/100ns
No Dimension board or other upgrades.
Verbose POST
...really a plain Cube overall.
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 14.4 seconds
From Drive Ready to Login Window: 76.3 seconds
Overall from bonk at POST to Login Window: 90.7 seconds
From Login Window to Workspace loaded: 25.5 seconds
Launching
Return to Login Window from Workspace: 7.5 seconds
Analysis
The overall startup process was an extra 10 seconds longer when factoring in the POST to drive ready part of the startup, however if we just look at drive ready to Login Window, it's about 4 seconds shorter than the entire Internal or External BlueSCSI startup on the same machine.
Launching applications were nearly the same, with the exception of one - Concurrence, which overall was 7 seconds longer than BlueSCSI.
When we consider that these are times with a mechanical hard drive and not solid state technology, the times are not unacceptable. Just think, platter drives with 9ms access times were fast and we just 'sat and waited' for things to complete.
I consider it somewhat impressive that the hard drive kept up in most things against new technology. Remembering the fact that the drive chosen for these tests, a drive that would be found in a DEC workstation computer, 1990 technology held its own, in its own respect. My 'new' money is on replacement technology, even if the past technology is still nice to see working, after all this time.
25MHz 68040
16MB RAM/100ns
No Dimension board or other upgrades.
Verbose POST
...really a plain Cube overall.
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 14.4 seconds
From Drive Ready to Login Window: 76.3 seconds
Overall from bonk at POST to Login Window: 90.7 seconds
From Login Window to Workspace loaded: 25.5 seconds
Launching
WordPerfect 1.0.1: 14.2 seconds
FrameMaker 3.2: 13.9 seconds
Concurrence: 2.77: 20.2 seconds
Diagram 2.5: 16.8 seconds
OpenWrite 2.1-8: 14.6 seconds
ParaSheet 1.7: 20.0 seconds
Return to Login Window from Workspace: 7.5 seconds
Analysis
The overall startup process was an extra 10 seconds longer when factoring in the POST to drive ready part of the startup, however if we just look at drive ready to Login Window, it's about 4 seconds shorter than the entire Internal or External BlueSCSI startup on the same machine.
Launching applications were nearly the same, with the exception of one - Concurrence, which overall was 7 seconds longer than BlueSCSI.
When we consider that these are times with a mechanical hard drive and not solid state technology, the times are not unacceptable. Just think, platter drives with 9ms access times were fast and we just 'sat and waited' for things to complete.
I consider it somewhat impressive that the hard drive kept up in most things against new technology. Remembering the fact that the drive chosen for these tests, a drive that would be found in a DEC workstation computer, 1990 technology held its own, in its own respect. My 'new' money is on replacement technology, even if the past technology is still nice to see working, after all this time.
Title: Re: BlueSCSI v2 Time Trials
Post by: Rob Blessin Black Hole on February 28, 2024, 01:59:32 AM
Post by: Rob Blessin Black Hole on February 28, 2024, 01:59:32 AM
Hello Cool: Nice work , I'm wondering if the SD cards are being throttled by our configs .... and if there is an optimized NeXTSTEP or Openstep config for blue scsi.
I guess what I'm getting at is with original SCSI Hard drives there are 5400, 7200, 10000 and even 15000 rpm drives along with 512K and 1024k block sizes also NeXTSTEP 2gb and with Openstep 4gb single partitions.
Also from my understanding the bus speeds vary between the earlier NeXT 25Mhz (better quality components)and later NeXT 33Mhz turbo hardware (bus runs at 66Mhz but most cutting components) as well as the onboard scsi controllers paired with optimized ram , some of those drives like the Seagate Barracuda 7200 RPM ran hot as well temperature wise but we no longer have that issue with the PCB scsi cards :)
I've also read 70NS is the factory spec but does parity or non parity ram make a difference as the manufactured / engineered ram chips newer are 2 or 3 chip (usually harvested from old 16Mb ram and set up on a new 4Mb chip template) and others have original factory 8 or 9 chips. Even the design of the chips may affect through put (Brian ad Dave looking at you lol)
Are Ceramic Caps faster than electrolytic?
If a Custom disk tab for the ZULU or Blue SCSI optimizing performance say emulating a Seagate Cheetah 10,000 rpm drive with 512K or 1024K blocks , this may configurable easier in the earlier SCSI2SD as it has the customizable XML file.... so in effect can we spoof more speed and performance out of the blue scsi or Zulu SCSI or SCSI2SD devices by spoofing an optimized or maximized drive ...stay tuned as I just thought of this lol.
It like we have vintage indy car and want to burn rubber as much as possible lol with out blowing shit up.
Clock speed doubler for the processor etc...
I know there are a few NeXTSTEP or Openstep software performance enhancing tips as evidenced by the Lighthouse apps launching faster in "Cools" benches I suspect custom api's :)
Cool, which bench mark software are you using ? And Yes I'll make you an excellent deal on a color station motherboard with ram etc ...
Ive heard that Swiss Bank actually developed a custom NeXTSTEP kernel , have never seen it for NeXT hardware to do just that .... is it just urban legend of larger hard drive support and support for more ram?
I actually own a 50 Mhz Turbo Color that has me scratching my head lol with a custom after market daughter board Ill make a video as it is very cool:)
I guess what I'm getting at is with original SCSI Hard drives there are 5400, 7200, 10000 and even 15000 rpm drives along with 512K and 1024k block sizes also NeXTSTEP 2gb and with Openstep 4gb single partitions.
Also from my understanding the bus speeds vary between the earlier NeXT 25Mhz (better quality components)and later NeXT 33Mhz turbo hardware (bus runs at 66Mhz but most cutting components) as well as the onboard scsi controllers paired with optimized ram , some of those drives like the Seagate Barracuda 7200 RPM ran hot as well temperature wise but we no longer have that issue with the PCB scsi cards :)
I've also read 70NS is the factory spec but does parity or non parity ram make a difference as the manufactured / engineered ram chips newer are 2 or 3 chip (usually harvested from old 16Mb ram and set up on a new 4Mb chip template) and others have original factory 8 or 9 chips. Even the design of the chips may affect through put (Brian ad Dave looking at you lol)
Are Ceramic Caps faster than electrolytic?
If a Custom disk tab for the ZULU or Blue SCSI optimizing performance say emulating a Seagate Cheetah 10,000 rpm drive with 512K or 1024K blocks , this may configurable easier in the earlier SCSI2SD as it has the customizable XML file.... so in effect can we spoof more speed and performance out of the blue scsi or Zulu SCSI or SCSI2SD devices by spoofing an optimized or maximized drive ...stay tuned as I just thought of this lol.
It like we have vintage indy car and want to burn rubber as much as possible lol with out blowing shit up.
Clock speed doubler for the processor etc...
I know there are a few NeXTSTEP or Openstep software performance enhancing tips as evidenced by the Lighthouse apps launching faster in "Cools" benches I suspect custom api's :)
Cool, which bench mark software are you using ? And Yes I'll make you an excellent deal on a color station motherboard with ram etc ...
Ive heard that Swiss Bank actually developed a custom NeXTSTEP kernel , have never seen it for NeXT hardware to do just that .... is it just urban legend of larger hard drive support and support for more ram?
I actually own a 50 Mhz Turbo Color that has me scratching my head lol with a custom after market daughter board Ill make a video as it is very cool:)
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 02, 2024, 01:37:25 AM
Post by: cooltr6 on March 02, 2024, 01:37:25 AM
@Rob Blessin Black Hole Appreciate the compliment! I'm sure there are a lot of things going on between the 'old' and 'new' technology.
For the configuration, except for Drive Geometry [the magic numbers of 139 and 4], my BlueSCSI config is setup as:
You do have a point discussing different drive RPM's and the changes in hardware over the time, along with memory and all other things that may influence performance one way or another. I'll just say that except for the Cube you integrated for me back in 2019 [aka 'The Loaded Cube']...all the others are still sitting on their original caps. As for the memory part you mentioned, I'd love to know myself if that makes any difference in things.
The thought of adding some 'extra spice' [not the extra sauce that was added in other ways] of getting more performance out of either BlueSCSI or ZuluSCSI...I'm game to try whatever. If we can squeeze even a second or two savings off of overall times...winner for us!
I'm not using any benchmark software, just a good ol' stopwatch to time the activities I'm performing for these tests. =)
For the configuration, except for Drive Geometry [the magic numbers of 139 and 4], my BlueSCSI config is setup as:
EnableUnitAttention=1 ; On
EnableSCSI2=1 ; On
MaxSyncSpeed=10
Vendor=BlueSCSI
Version=2
SectorsPerTrack=139
HeadsPerCylinder=4
You do have a point discussing different drive RPM's and the changes in hardware over the time, along with memory and all other things that may influence performance one way or another. I'll just say that except for the Cube you integrated for me back in 2019 [aka 'The Loaded Cube']...all the others are still sitting on their original caps. As for the memory part you mentioned, I'd love to know myself if that makes any difference in things.
The thought of adding some 'extra spice' [not the extra sauce that was added in other ways] of getting more performance out of either BlueSCSI or ZuluSCSI...I'm game to try whatever. If we can squeeze even a second or two savings off of overall times...winner for us!
I'm not using any benchmark software, just a good ol' stopwatch to time the activities I'm performing for these tests. =)
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 02, 2024, 01:41:22 AM
Post by: cooltr6 on March 02, 2024, 01:41:22 AM
BlueSCSI Time Trials - The 'Loaded Cube' and a Standard Hard Drive
25MHz 68040
64MB RAM/100ns
Dimension Board with 64MB RAM - NTSC
Has an internal CD-ROM
Verbose POST
Note: For the purposes of this test, the CD-ROM drive was disconnected, as it adds time to the overall startup process. Since we're looking for from bonk at POST to Login Window, we want to make this test as closely configured as other the other systems used.
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 21.9 seconds
From Drive Ready to Login Window: 56.1 seconds
+ 11.0 seconds for the Dimension Board to come available.
Overall from bonk at POST to Login Window: 78.0 seconds.
89.0 seconds (including Dimension Board]
From Login Window to Workspace loaded: 22.5 seconds
Launching
Return to Login Window from Workspace: 4.9 seconds
Analysis
As noted in the other test for the 'Loaded Cube', memory does make a difference, however with that said, the addition of the Dimension Board did add an extra 11 seconds to the overall startup time. Expected, but still something just to be aware of.
The 'Loaded Cube' throughout all of the platter drive testing was the slowest to get from Bonk at POST to Drive Ready. This test was ran twice, with the same time being reported. Even with that, from Drive Ready to Login window, the 'Loaded Cube' performed better than the 'Not-so-Loaded' Cube and was 2 seconds better performing than the Turbo Mono Slab.
Do keep in mind the 'Loaded Cube' also has a Dimension Board with 64MB RAM installed, so you have to add in an extra 11 seconds. The Overall bonk at POST to Login Window was just 5 seconds longer, before adding in the 11 seconds for the Dimension Board.
Most applications launched in the same times as the 'Not-so-Loaded' Cube with the platter drive. There were a couple though that were slower — WordPerfect and OpenWrite at 4 and 6 seconds slower respectively.
Return to Login Window from Workspace for this setup reported the fastest time so far of 4.9 seconds, which is nearly 3 seconds faster than the 'Not-so-Loaded' Cube and just 1 second faster than the Mono Turbo Slab with platters.
Still gotta hand it to the platter drive...at nearly 30 years old, still holding its own against new technology.
25MHz 68040
64MB RAM/100ns
Dimension Board with 64MB RAM - NTSC
Has an internal CD-ROM
Verbose POST
Note: For the purposes of this test, the CD-ROM drive was disconnected, as it adds time to the overall startup process. Since we're looking for from bonk at POST to Login Window, we want to make this test as closely configured as other the other systems used.
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 21.9 seconds
From Drive Ready to Login Window: 56.1 seconds
+ 11.0 seconds for the Dimension Board to come available.
Overall from bonk at POST to Login Window: 78.0 seconds.
89.0 seconds (including Dimension Board]
From Login Window to Workspace loaded: 22.5 seconds
Launching
WordPerfect 1.0.1: 18.4 seconds
FrameMaker 3.2: 12.6 seconds
Concurrence: 2.77: 15.2 seconds
Diagram 2.5: 16.4 seconds
OpenWrite 2.1-8: 20.0 seconds
ParaSheet 1.7: 19.9 seconds
Return to Login Window from Workspace: 4.9 seconds
Analysis
As noted in the other test for the 'Loaded Cube', memory does make a difference, however with that said, the addition of the Dimension Board did add an extra 11 seconds to the overall startup time. Expected, but still something just to be aware of.
The 'Loaded Cube' throughout all of the platter drive testing was the slowest to get from Bonk at POST to Drive Ready. This test was ran twice, with the same time being reported. Even with that, from Drive Ready to Login window, the 'Loaded Cube' performed better than the 'Not-so-Loaded' Cube and was 2 seconds better performing than the Turbo Mono Slab.
Do keep in mind the 'Loaded Cube' also has a Dimension Board with 64MB RAM installed, so you have to add in an extra 11 seconds. The Overall bonk at POST to Login Window was just 5 seconds longer, before adding in the 11 seconds for the Dimension Board.
Most applications launched in the same times as the 'Not-so-Loaded' Cube with the platter drive. There were a couple though that were slower — WordPerfect and OpenWrite at 4 and 6 seconds slower respectively.
Return to Login Window from Workspace for this setup reported the fastest time so far of 4.9 seconds, which is nearly 3 seconds faster than the 'Not-so-Loaded' Cube and just 1 second faster than the Mono Turbo Slab with platters.
Still gotta hand it to the platter drive...at nearly 30 years old, still holding its own against new technology.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 02, 2024, 01:45:44 AM
Post by: cooltr6 on March 02, 2024, 01:45:44 AM
BlueSCSI Time Trials - The 'Loaded Cube'
25MHz 68040
64MB RAM/100ns
Dimension Board with 64MB RAM - NTSC
Has an internal CD-ROM
Verbose POST
Note: For the purposes of this test, the CD-ROM drive was disconnected, as it adds time to the overall startup process. Since we're looking for from bonk at POST to Login Window, we want to make this test as closely configured as other the other systems used.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 64.6 seconds
+ 11.0 seconds for the Dimension Board to come available.
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 22.5 seconds
Launching
Return to Login Window from Workspace: 6.2 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 68.8 seconds
+ 11.0 seconds for the Dimension Board to come available.
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 45.9 seconds
Launching
Return to Login Window from Workspace: 6.2 seconds
Analysis
To start with, extra memory does make a difference. In both tests with the Internal and External BlueSCSI, it more than edged out the 'Not-so-Loaded' Cube by almost 20 seconds in the Bonk at POST to Login Window test.. Workspace and CD mounting time across both Cubes were right about the same times, differing by 1-3 seconds between Internal and External versions.
There were minor differences in launching applications between the two Cubes and the Internal and External BlueSCSI devices. The most striking was opening OpenWrite with the External BlueSCSI on the Loaded cube. It took 8 seconds longer than it did on the 'Not-so-Loaded' Cube. Otherwise, application launching times were within acceptable limits.
Workspace to Login Screen were exactly the same at 6.2 seconds.
25MHz 68040
64MB RAM/100ns
Dimension Board with 64MB RAM - NTSC
Has an internal CD-ROM
Verbose POST
Note: For the purposes of this test, the CD-ROM drive was disconnected, as it adds time to the overall startup process. Since we're looking for from bonk at POST to Login Window, we want to make this test as closely configured as other the other systems used.
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 64.6 seconds
+ 11.0 seconds for the Dimension Board to come available.
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 22.5 seconds
Launching
WordPerfect 1.0.1: 10.1 seconds
FrameMaker 3.2: 10.6 seconds
Concurrence: 2.77: 12.8 seconds
Diagram 2.5: 15.3 seconds
OpenWrite 2.1-8: 10.3 seconds
ParaSheet 1.7: 19.0 seconds
Return to Login Window from Workspace: 6.2 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 68.8 seconds
+ 11.0 seconds for the Dimension Board to come available.
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 45.9 seconds
Launching
WordPerfect 1.0.1: 9.6 seconds
FrameMaker 3.2: 11.1 seconds
Concurrence: 2.77: 12.5 seconds
Diagram 2.5: 15.4 seconds
OpenWrite 2.1-8: 19.1 seconds
ParaSheet 1.7: 19.3 seconds
Return to Login Window from Workspace: 6.2 seconds
Analysis
To start with, extra memory does make a difference. In both tests with the Internal and External BlueSCSI, it more than edged out the 'Not-so-Loaded' Cube by almost 20 seconds in the Bonk at POST to Login Window test.. Workspace and CD mounting time across both Cubes were right about the same times, differing by 1-3 seconds between Internal and External versions.
There were minor differences in launching applications between the two Cubes and the Internal and External BlueSCSI devices. The most striking was opening OpenWrite with the External BlueSCSI on the Loaded cube. It took 8 seconds longer than it did on the 'Not-so-Loaded' Cube. Otherwise, application launching times were within acceptable limits.
Workspace to Login Screen were exactly the same at 6.2 seconds.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 02, 2024, 01:55:18 AM
Post by: cooltr6 on March 02, 2024, 01:55:18 AM
Current state of testing...
Things are going to be on hold for a little bit. The Mono Slab [non-Turbo] is on the bench, taken apart and all prepared for a recap job. Gonna be my first attempt at it...wish me luck!
Also, for whatever reason...including some very late night stupidity from me...my SCSI2SD card no longer is working as expected. The Mangler struck again. Gotta get with Rob to see about having an SD Card re-imaged and figure out why the SCSI2SD card won't see all drives.
With that, I wanted to post a table of results. They're all in the previous postings, but a scorecard 'up until now' will be good for folks to look at rather than having to read through all the posts just to get an overall picture.
Once the Mono Slab is back up and running, it'll get the same treatment of platter, Internal and External BlueSCSI tests.
THEN after I get my SCSI2SD card working again...everyone will get a going over with that and the testing will be done.
For those who have taken time to read the postings in this thread, my many thanks. More to come as soon as stuff is off the bench and back on the happy working side of things.
Things are going to be on hold for a little bit. The Mono Slab [non-Turbo] is on the bench, taken apart and all prepared for a recap job. Gonna be my first attempt at it...wish me luck!
Also, for whatever reason...including some very late night stupidity from me...my SCSI2SD card no longer is working as expected. The Mangler struck again. Gotta get with Rob to see about having an SD Card re-imaged and figure out why the SCSI2SD card won't see all drives.
With that, I wanted to post a table of results. They're all in the previous postings, but a scorecard 'up until now' will be good for folks to look at rather than having to read through all the posts just to get an overall picture.
Once the Mono Slab is back up and running, it'll get the same treatment of platter, Internal and External BlueSCSI tests.
THEN after I get my SCSI2SD card working again...everyone will get a going over with that and the testing will be done.
For those who have taken time to read the postings in this thread, my many thanks. More to come as soon as stuff is off the bench and back on the happy working side of things.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 04, 2024, 12:08:30 AM
Post by: cooltr6 on March 04, 2024, 12:08:30 AM
We're back testing. The recap project for the Mono Slab was a success! First ever recap job, took almost 3 hours. Taking time, dealing with a couple of traces that came apart and a few other little unforeseen corners...the Mono Slab came off the operating table and made a full recovery.
Because of that success, I ended up tackling the Turbo Mono Slab and the other Cube. Now all the Black hardware in the stable are sitting on new caps. Many thanks to @barcher174 for putting together the comprehensive recap details.
Now, on to the test results for the Mono Slab.
Because of that success, I ended up tackling the Turbo Mono Slab and the other Cube. Now all the Black hardware in the stable are sitting on new caps. Many thanks to @barcher174 for putting together the comprehensive recap details.
Now, on to the test results for the Mono Slab.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 04, 2024, 12:11:18 AM
Post by: cooltr6 on March 04, 2024, 12:11:18 AM
BlueSCSI Time Trials - NeXTstation Mono and a Standard Hard Drive
25MHz 68040
16MB RAM/100ns
No other upgrades.
Verbose POST
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 21.2 seconds
From Drive Ready to Login Window: 68.9 seconds
Overall from bonk at POST to Login Window: 90.1 seconds
From Login Window to Workspace loaded: 24.8 seconds
Return to Login Window from Workspace: 7.8 seconds
Launching
Analysis
Not too surprising at the numbers put up by the platter drive here. When comparing against the two Cubes at 25MHz, performance almost matched the Loaded Cube, with not much variation in times.
Across the board, expected performance from the platter drive. Not much else to say here, except what I've said before...when we consider this is what we used to deal with on a daily basis, it's still rock solid.
25MHz 68040
16MB RAM/100ns
No other upgrades.
Verbose POST
DEC RZ26L 1GB SCSI Hard Drive [SCSI-2, 5400± rpm, avg. seek time 9.5ms] with a full install of OPENSTEP 4.2 [including Developer Tools]. There is only one partition on this drive. No mountable CD's. This differs from the BlueSCSI setup.
This computer can connect to the Internet and the local NFS share.
Test Results
Platter Drive
From bonk at POST to Drive Ready: 21.2 seconds
From Drive Ready to Login Window: 68.9 seconds
Overall from bonk at POST to Login Window: 90.1 seconds
From Login Window to Workspace loaded: 24.8 seconds
Return to Login Window from Workspace: 7.8 seconds
Launching
WordPerfect 1.0.1: 19.9 seconds
FrameMaker 3.2: 15.2 seconds
Concurrence: 2.77: 16.9 seconds
Diagram 2.5: 16.5 seconds
OpenWrite 2.1-8: 12.4 seconds
ParaSheet 1.7: 20.5 seconds
Analysis
Not too surprising at the numbers put up by the platter drive here. When comparing against the two Cubes at 25MHz, performance almost matched the Loaded Cube, with not much variation in times.
Across the board, expected performance from the platter drive. Not much else to say here, except what I've said before...when we consider this is what we used to deal with on a daily basis, it's still rock solid.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 04, 2024, 12:13:39 AM
Post by: cooltr6 on March 04, 2024, 12:13:39 AM
BlueSCSI Time Trials - NeXTstation Mono
25MHz 68040
16MB RAM/100ns
No other upgrades.
Verbose POST
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 78.4 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 23.5 seconds
Launching
Return to Login Window from Workspace: 7.9 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 79.3 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 49.1 seconds
Launching
Return to Login Window from Workspace: 6.8 seconds
Analysis
Testing the Mono Slab with BlueSCSI, there were some interesting things that came up. Overall, both Internal and External BlueSCSI performed a bit better than the Not-so-Loaded Cube, but didn't match the overall performance of the Loaded Cube. Taking into account the Loaded Cube has a Dimension Board installed, that added an extra 11 seconds to the startup process.
Internal BlueSCSI though did stand the test against the other 25MHz machines, matching the Not-so-Loaded Cube. External BlueSCSI performed about the same as it did for others in the 25MHz class.
However, there were some surprises in application launching, both for Internal and External BlueSCSI. WordPerfect launched almost 2 seconds faster on the External version, when compared to Internal. Others were almost a second faster, with External. Even Return to Login Window from Workspace, was faster with External, a second faster than both Internal and platter..
Let's add another wrinkle into this mix. Just before doing this testing, I had completed the recap project for this unit. There were some touch and go moments, thanks to a little trace rot, but thankfully, those were fixed and there was life in this little gem.
25MHz 68040
16MB RAM/100ns
No other upgrades.
Verbose POST
BlueSCSI v2 [internal and external], using a 32GB SanDisk High Endurance Card [100MB/s read, 40MB/s write] and a full install of OPENSTEP 4.2 [including Developer Tools]
There are 3-2GB drives setup at SCSI ID's 1, 2 and 3. CD5 [SCSI ID 5] is loaded with 5 Peanuts CD's. Tests were performed with the same SD card, but using the internal and external versions.
Since I'm using the same card as the Mono Turbo Slab, this computer can connect to the Internet and the local NFS share.
Test Results
Internal BlueSCSI v2
From bonk at POST to Login Window: 78.4 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 23.5 seconds
Launching
WordPerfect 1.0.1: 11.1 seconds
FrameMaker 3.2: 12.7 seconds
Concurrence: 2.77: 13.2 seconds
Diagram 2.5: 15.3 seconds
OpenWrite 2.1-8: 9.2 seconds
ParaSheet 1.7: 19.4 seconds
Return to Login Window from Workspace: 7.9 seconds
External BlueSCSI v2
From bonk at POST to Login Window: 79.3 seconds
From Login Window to Workspace loaded [including mounting the first CD - Peanuts 1]: 49.1 seconds
Launching
WordPerfect 1.0.1: 9.8 seconds
FrameMaker 3.2: 11.8 seconds
Concurrence: 2.77: 13.7 seconds
Diagram 2.5: 14.8 seconds
OpenWrite 2.1-8: 10.5 seconds
ParaSheet 1.7: 18.4 seconds
Return to Login Window from Workspace: 6.8 seconds
Analysis
Testing the Mono Slab with BlueSCSI, there were some interesting things that came up. Overall, both Internal and External BlueSCSI performed a bit better than the Not-so-Loaded Cube, but didn't match the overall performance of the Loaded Cube. Taking into account the Loaded Cube has a Dimension Board installed, that added an extra 11 seconds to the startup process.
Internal BlueSCSI though did stand the test against the other 25MHz machines, matching the Not-so-Loaded Cube. External BlueSCSI performed about the same as it did for others in the 25MHz class.
However, there were some surprises in application launching, both for Internal and External BlueSCSI. WordPerfect launched almost 2 seconds faster on the External version, when compared to Internal. Others were almost a second faster, with External. Even Return to Login Window from Workspace, was faster with External, a second faster than both Internal and platter..
Let's add another wrinkle into this mix. Just before doing this testing, I had completed the recap project for this unit. There were some touch and go moments, thanks to a little trace rot, but thankfully, those were fixed and there was life in this little gem.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 04, 2024, 12:42:27 AM
Post by: cooltr6 on March 04, 2024, 12:42:27 AM
Current State of Testing and Some Thoughts
Attached is the current table of results from Platter and BlueSCSI testing. I got in contact with Rob Blessin, who is going to help me get my ailing SCSI2HD card hopefully back amongst the land of the living. Once that happens, the final section of testing can be completed. Soon as things are ready, I'll start posting results.
Now for some thoughts...
The opinions and results from BlueSCSI testing are mine and mine alone. My bringing this work log was not in any way to promote or sell this product. I have purchased for myself, in total, 12 BlueSCSI devices [2 external, 10 internal] and was not asked by the seller I purchased from to plug or otherwise advertise.
When looking for viable replacements for not only Black hardware, but for the vintage Apple II and Macintosh hardware in my stable, I weighted out the pros and cons of multiple devices and came to the decision for BlueSCSI...for a number of reasons [to list just some of those]:
1. Ease of setup. Create an empty drive image, name it, copy it to the SD card and go. The same SD-card formatting works whether it's a NeXT, Apple IIgs or a Mac. Also, Drive Geometry setup is easier.
2. Ease of making multiple 'disks'. Because I'm using this in Black hardware as well as the other SCSI-capable machines in my lineup, at least for NeXT, it's easy. Create one GM version of all applications, network setup and the sort...copy the drive files off the SD card, load up 3 cards with the drive files...done. After that, it's just a bit of housekeeping to change hostnames and IP addresses. For the Macs it'd be the same thing.
3. Being able to mount CD-ROM images. This was a thing I had been looking for. At the time I had by SCSI2SD v5.1 card, there was no support for such a thing. Once I found that there was a way with BlueSCSI to drop a bunch of CD-ROM images to the SD card, that slid the balance solely on the side of BlueSCSI. This works happily for NeXT and the Mac side of things. Haven't tried it with the IIgs yet.
4. Use of a Shared Folder to get items back and forth between Vintage [Macs] and Modern hardware. This was another thing that was not available with SCSI2SD v5.1. Although still 'beta' with BlueSCSI, it does work quite well, on the Mac side of things. This wouldn't be available for NeXT, unless someone writes a NeXT-specific version of the transfer software, nor for the IIgs.
5. Getting the Macs on the network. For the Macintoshes that are going to be online, the WiFi enabled version of Internal BlueSCSI fit my needs. I can't see paying exorbitant prices for Farallon PhoneNet boxes to get these things networked up. The WiFi is just as fast as original LocalTalk 'hardwired'...so, get a two-for-one special...networking and SCSI all wrapped into one package.
6. Finding a compact package that'd fit neatly into an Apple IIgs. Among many of the machines I have, there are 3 Apple IIgs's in the mix...one with a still functioning platter drive/power supply, the other two with nothing. The Woz Limited Edition remains bone stock, no upgrades, no additions. The other, that had nothing, I purchased a SCSI card, which is a near clone of the Apple SCSI card, which fetches a pretty penny today. The external version, once tied to the internal ribbon cable fit inside the case. Performance for that almost matches that of the platter drive that's in the other Apple IIgs, so it fit the bill...literally.
Bringing the results of this testing has not only been a fun learning experience, but one that I've enjoyed immensely. Arming people with information as they make the same decisions as I did...gotta replace failing drives, what kind of device to use...since there are a number of them out there in the wild. Some may offer performance advantages or ease of setup over other things, that's going to be something that each individual will need to look at and come to their own decision.
Continued thanks as you read through this lengthy work log.
As noted in an earlier post, soon as the SCSI2SD card is back up and running, I'll finish of the last round of testing.
P.S. I know I mixed some non-NeXT related stuff in this post...but the Macs and IIgs's that are here..deserve mention. =)
Attached is the current table of results from Platter and BlueSCSI testing. I got in contact with Rob Blessin, who is going to help me get my ailing SCSI2HD card hopefully back amongst the land of the living. Once that happens, the final section of testing can be completed. Soon as things are ready, I'll start posting results.
Now for some thoughts...
The opinions and results from BlueSCSI testing are mine and mine alone. My bringing this work log was not in any way to promote or sell this product. I have purchased for myself, in total, 12 BlueSCSI devices [2 external, 10 internal] and was not asked by the seller I purchased from to plug or otherwise advertise.
When looking for viable replacements for not only Black hardware, but for the vintage Apple II and Macintosh hardware in my stable, I weighted out the pros and cons of multiple devices and came to the decision for BlueSCSI...for a number of reasons [to list just some of those]:
1. Ease of setup. Create an empty drive image, name it, copy it to the SD card and go. The same SD-card formatting works whether it's a NeXT, Apple IIgs or a Mac. Also, Drive Geometry setup is easier.
2. Ease of making multiple 'disks'. Because I'm using this in Black hardware as well as the other SCSI-capable machines in my lineup, at least for NeXT, it's easy. Create one GM version of all applications, network setup and the sort...copy the drive files off the SD card, load up 3 cards with the drive files...done. After that, it's just a bit of housekeeping to change hostnames and IP addresses. For the Macs it'd be the same thing.
3. Being able to mount CD-ROM images. This was a thing I had been looking for. At the time I had by SCSI2SD v5.1 card, there was no support for such a thing. Once I found that there was a way with BlueSCSI to drop a bunch of CD-ROM images to the SD card, that slid the balance solely on the side of BlueSCSI. This works happily for NeXT and the Mac side of things. Haven't tried it with the IIgs yet.
4. Use of a Shared Folder to get items back and forth between Vintage [Macs] and Modern hardware. This was another thing that was not available with SCSI2SD v5.1. Although still 'beta' with BlueSCSI, it does work quite well, on the Mac side of things. This wouldn't be available for NeXT, unless someone writes a NeXT-specific version of the transfer software, nor for the IIgs.
5. Getting the Macs on the network. For the Macintoshes that are going to be online, the WiFi enabled version of Internal BlueSCSI fit my needs. I can't see paying exorbitant prices for Farallon PhoneNet boxes to get these things networked up. The WiFi is just as fast as original LocalTalk 'hardwired'...so, get a two-for-one special...networking and SCSI all wrapped into one package.
6. Finding a compact package that'd fit neatly into an Apple IIgs. Among many of the machines I have, there are 3 Apple IIgs's in the mix...one with a still functioning platter drive/power supply, the other two with nothing. The Woz Limited Edition remains bone stock, no upgrades, no additions. The other, that had nothing, I purchased a SCSI card, which is a near clone of the Apple SCSI card, which fetches a pretty penny today. The external version, once tied to the internal ribbon cable fit inside the case. Performance for that almost matches that of the platter drive that's in the other Apple IIgs, so it fit the bill...literally.
Bringing the results of this testing has not only been a fun learning experience, but one that I've enjoyed immensely. Arming people with information as they make the same decisions as I did...gotta replace failing drives, what kind of device to use...since there are a number of them out there in the wild. Some may offer performance advantages or ease of setup over other things, that's going to be something that each individual will need to look at and come to their own decision.
Continued thanks as you read through this lengthy work log.
As noted in an earlier post, soon as the SCSI2SD card is back up and running, I'll finish of the last round of testing.
P.S. I know I mixed some non-NeXT related stuff in this post...but the Macs and IIgs's that are here..deserve mention. =)
Title: Re: BlueSCSI v2 Time Trials
Post by: sappas on March 15, 2024, 03:18:06 PM
Post by: sappas on March 15, 2024, 03:18:06 PM
Hello, you could use Drive Performance, i've a zuluscsi 2. https://www.nextcomputers.org/NeXTfiles/Software/OPENSTEP/Apps/Benchmarks/DrivePerformance.1.1.NI.b.tar.gz
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 15, 2024, 07:07:19 PM
Post by: cooltr6 on March 15, 2024, 07:07:19 PM
Quote from: sappas on March 15, 2024, 03:18:06 PMHello, you could use Drive Performance, i've a zuluscsi 2. https://www.nextcomputers.org/NeXTfiles/Software/OPENSTEP/Apps/Benchmarks/DrivePerformance.1.1.NI.b.tar.gz
@sappas Thank you for that! I'll certainly go back and get some drive performance times. Most appreciated.
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 16, 2024, 04:05:08 AM
Post by: cooltr6 on March 16, 2024, 04:05:08 AM
Greetings folks. Back again with some more detail. Sappas let us know about a drive testing application called DrivePerformance. I started with some testing using this application and took some detail from the "List of Drives" info panel from DrivePerformance.
I performed tests with the two NeXTstations just a bit ago...I'll get the Cubes out later on today or tomorrow and post those scores when they're ready. Attached is a screenshot of the numbers thus far.
Please note: The list of drives found in DrivePerformance -- I have no idea of what NeXT computer was used to test these drives...just putting stuff there for your information. =)
I performed tests with the two NeXTstations just a bit ago...I'll get the Cubes out later on today or tomorrow and post those scores when they're ready. Attached is a screenshot of the numbers thus far.
Please note: The list of drives found in DrivePerformance -- I have no idea of what NeXT computer was used to test these drives...just putting stuff there for your information. =)
Title: Re: BlueSCSI v2 Time Trials
Post by: cooltr6 on March 16, 2024, 04:42:40 PM
Post by: cooltr6 on March 16, 2024, 04:42:40 PM
Filled in some more blanks. Got the Cubes out and tested things.
Quite a surprise to see how the Cubes stacked up against the Slabs. The most surprising is the Loaded Cube in Read tests with BlueSCSI at almost .5MB faster than the Turbo Slab. In looking at Overall Indexes, the Loaded Cube edged out the Mono Turbo Slab by a mere .2 of a point overall in all tests performed. The Not-so-Loaded Cube and the Mono Slab, were nearly equal to each other.
Hard drive performance across the board is not too surprising, but still did put up a good fight. Highest overall index at 1.32 with the Loaded Cube and the lowest with the Not-so-Loaded Cube at 0.89. The Slabs had the same variation in overall indexes, but only .1 of a point between the two.
For the most part, we'd expect SD-card technology to win out over platter drives. I'm going to be interested in seeing the results of SCSI2SD in the same test setup. I really need to get that card off to Rob, so it can be put on the operating table for an exploratory.
Attached are the results to date, with both NeXTstation and Cube tests. We're nearing the end, all that's left is SCSI2SD detail.
Quite a surprise to see how the Cubes stacked up against the Slabs. The most surprising is the Loaded Cube in Read tests with BlueSCSI at almost .5MB faster than the Turbo Slab. In looking at Overall Indexes, the Loaded Cube edged out the Mono Turbo Slab by a mere .2 of a point overall in all tests performed. The Not-so-Loaded Cube and the Mono Slab, were nearly equal to each other.
Hard drive performance across the board is not too surprising, but still did put up a good fight. Highest overall index at 1.32 with the Loaded Cube and the lowest with the Not-so-Loaded Cube at 0.89. The Slabs had the same variation in overall indexes, but only .1 of a point between the two.
For the most part, we'd expect SD-card technology to win out over platter drives. I'm going to be interested in seeing the results of SCSI2SD in the same test setup. I really need to get that card off to Rob, so it can be put on the operating table for an exploratory.
Attached are the results to date, with both NeXTstation and Cube tests. We're nearing the end, all that's left is SCSI2SD detail.