Re: Sun SS10 model 30 Performance
Date: 24 Oct 92 13:14:48 GMT
Message-ID: <1992Oct24.131448.1921_at_ukw.uucp>
jjg_at_pt.com (John Grana) writes:
>I think that the 1.7MB/Sec is pretty good (on writes). The disk drive
>can probably only write the actual "spinning media" at around 2.1 or so MB/sec.
>Also, notice how even small I/O requests (512 to 2K) go pretty fast.
>I have been doing lots of SCSI benchmarking the past few days on an SS10
>and have been pretty happy ('course I would since our SBus H.A. (SBS430)
>is getting around 2 MB/Sec!) For a reference, run the IOZONE benchmark
>on a SS2. You might be suprised, esp. at small requests (300-500KB/sec).
>It seems the old sweet spot at 8K is now spread over most all I/O sizes
>(at least on the SS10 running 4.1.3).
For comparison, here are a few Bonnie results for my SS IPX, on-board host adapter, various disks, all at the end of a relatively long SCSI chain. (But all ran synchronously, so it shouldn't matter where they were located.) All disks had a file system. The DSP3105 is my system disk; I used a partition that is at the "end" of the disk, while all others had a single partition. All partitions except those on the M2624 were empty at the time of the test. I got another DSP3105 now, so I will repeat the run with a single, empty partition. When I got the results, I'll post a follow-up.
-------Sequential Output-------- ---Sequential Input-- --Random--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
DSP3105 100 889 99.7 1691 29.2 500 19.2 664 93.5 2675 59.9 60.5 12.2 IBM 0663 100 841 97.4 1609 28.6 401 16.5 645 92.4 2282 53.6 55.8 12.3 M2623FA 100 845 96.5 1634 30.1 418 17.0 574 87.6 1805 40.1 49.5 11.3 M2624SA 100 870 98.0 1429 24.9 442 17.0 639 90.4 1835 39.4 45.2 9.9 SiliconD 50 807 98.9 3154 77.3 1733 79.2 662 95.7 3717 92.9 310.2 67.8
The last disk is a SiliconDisk, a SCSI RAM disk with a 68020 and an Emulex (whatever, a SCSI Fast chip).
I have run a simpler test, something like 10 MBytes with dd from /dev/zero to a file system file, with the SiliconDisk on an SS 10/30. It ran at about 3.5 MBytes/sec, slow SCSI mode probably because of the cheap cable I used. The better throughput is certainly because of the faster CPU.
-- | ...!unido!ukw!lupe (German EUNet, "bang") | Disclaimer: | | lupe_at_ukw.UUCP (German EUNet, domain) | As I am self-employed, | | suninfo!alanya!lupe (Sun Germany) | this *is* the opinion | | Res non sunt complicanda praeter necessitatem. | of my employer. |Received on Sat Oct 24 1992 - 14:14:48 CET