Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 5 Dec 2001 14:44:08 -0600 (CST)
From:      Nick Rogness <nick@rogness.net>
To:        freebsd-scsi@freebsd.org
Cc:        mjacob@feral.com
Subject:   Qlogic 2100 FC and COMPAQ HSG60
Message-ID:  <Pine.BSF.4.21.0112051355280.84510-100000@cody.jharris.com>

next in thread | raw e-mail | index | archive | help

I'm having a couple of issues that I need some help with.  I have a Qlogic
2100 FC card connected to a Compaq HSG60 FC controller via a 8 port
switch.  The HSG60 has 10 18.2GB disks tied to it, configured in a RAID5
config.  This all resides on a 4.4-STABLE machine (sup'd yesterday).

Here's the relative info from dmesg:

isp0: <Qlogic ISP 2100 PCI FC-AL Adapter> port 0x1400-0x14ff mem
0x40000000-0x40000fff irq 9 at device 10.0 on pci2
...
pass0 at isp0 bus 0 target 129 lun 0
pass0: <DEC HSG60CCL V85L> Fixed Storage Arrray SCSI-2 device
pass0: 100.000MB/s transfers, Tagged Queueing Enabled
pass3 at isp0 bus 0 target 130 lun 0
pass3: <DEC HSG60CCL V85L> Fixed Storage Arrray SCSI-2 device
pass3: 100.000MB/s transfers, Tagged Queueing Enabled
da0 at isp0 bus 0 target 129 lun 1
da0: <DEC HSG60 V85L> Fixed Direct Access SCSI-2 device
da0: 100.000MB/s transfers, Tagged Queueing Enabled
da0: 103101MB (211152379 512 byte sectors: 255H 63S/T 13143C)


I have this disk mounted as /mnt.  I dumped / to /mnt to test it.

ninja# df
Filesystem  1K-blocks     Used    Avail Capacity  Mounted on
/dev/ad0s1a   1016303    41608   893391     4%    /
/dev/ad0s1f  12850409  3319226  8503151    28%    /usr
/dev/ad0s1e   4065262    96213  3643829     3%    /var
procfs              4        4        0   100%    /proc
/dev/da0s2e 102320350    41608 94093114     0%    /mnt

The dump seemed to take a long time.  So after the dump finished, I ran a
dump on /mnt to /dev/null to get some read performance data. I got some
weird results.  It takes several minutes to complete (~5 minutes), but
dump reports it only took seconds:

ninja# time dump 0af /dev/null /mnt
  DUMP: Date of this level 0 dump: Wed Dec  5 13:16:12 2001
  DUMP: Date of last level 0 dump: the epoch
  DUMP: Dumping /dev/da0s2e (/mnt) to /dev/null
  DUMP: mapping (Pass I) [regular files]
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated 51521 tape blocks.
  DUMP: dumping (Pass III) [directories]
  DUMP: dumping (Pass IV) [regular files]
  DUMP: DUMP: 48409 tape blocks on 1 volume
  DUMP: finished in 8 seconds, throughput 6051 KBytes/sec
  DUMP: Closing /dev/null
  DUMP: DUMP IS DONE
4.218u 7.059s 4:57.57 3.7%      310+12393k 0+2io 0pf+0w

After these results, I decided to try it with tar:

ninja# time tar -cPlf /mnt/root.tar /
tar: /dev/sa0.ctl: minor number too large; not dumped
0.016u 0.589s 0:03.52 16.7%     320+299k 2158+646io 0pf+0w


I also installed bonnie and tested:

ninja# cd /mnt
ninja# bonnie -s 1024

-------Sequential Output-------- ---Sequential Input-- --Random-- 
-Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU  /sec %CPU
20638 40.6 18917 15.3 12813 11.4 26054 87.8 38594 17.6 376.2  2.0

The bonnie and tar results seemed to look reasonable (I think).  

However, dumping 41MB should not take 5 minutes to finish (even though it
says 8 seconds).

So my question is: 

What does dump do differently that I'm not seeing in bonnie or tar?


Nick Rogness <nick@rogness.net>
 - Keep on Routing in a Free World...
  "FreeBSD: The Power to Serve!"




To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0112051355280.84510-100000>