Quinton Jansen wrote:
What speeds are others getting when using an LTO-3 drive?
If I'm reading the numbers correctly, 2GB in 13 minutes (10GB/hour) is way too
slow (should be around 200G/hour).
I've got a brand new Quantum Superloader-3, LTO3 attached to a dedicated SCSI
bus on a HP DL/380, running bacula 1.38.5, compiled from source.
************************************************************
tapeinfo -f /dev/sg0
Product Type: Tape Drive
Vendor ID: 'CERTANCE'
Product ID: 'ULTRIUM 3 '
Revision: '1794'
Attached Changer: No
SerialNumber: 'JD0084S '
MinBlock:1
MaxBlock:16777215
SCSI ID: 6
SCSI LUN: 0
Ready: yes
BufferedMode: yes
Medium Type: Not Loaded
Density Code: 0x44
BlockSize: 0
DataCompEnabled: yes
DataCompCapable: yes
DataDeCompEnabled: yes
CompType: 0x1
DeCompType: 0x1
Block Position: 34415
************************************************************
Feb 15 00:39:23 localhost kernel: SCSI subsystem initialized
Feb 15 00:39:23 localhost kernel: scsi0 : Adaptec AIC79XX PCI-X SCSI HBA
DRIVER, Rev 1.3.11
Feb 15 00:39:23 localhost kernel: <Adaptec 29320ALP Ultra320 SCSI
adapter>
Feb 15 00:39:23 localhost kernel: aic7901: Ultra320 Wide Channel A,
SCSI Id=7, PCI-X 67-100Mhz, 512 SCBs
Feb 15 00:39:23 localhost kernel: Type: Sequential-Access
ANSI SCSI revision: 04
Feb 15 00:39:23 localhost kernel: st 0:0:6:0: Attached scsi tape st0<4>st0:
try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575
Feb 15 00:39:23 localhost kernel: Type: Medium Changer
ANSI SCSI revision: 02
Feb 15 00:40:52 localhost kernel: st 0:0:6:0: Attached scsi generic sg0 type 1
Feb 15 00:40:52 localhost kernel: 0:0:6:1: Attached scsi generic sg1 type 8
Do you want to run the simplified test (s) with one tape
or the complete multiple tape (m) test: (s/m) s
Simple test (single tape) selected.
Wrote Volume label for volume "TestVolume1".
Wrote Start of Session label.
16:44:34 Begin writing Bacula records to tape ...
Wrote blk_block=5000, dev_blk_num=4999 VolBytes=322,495,472 rate=2733.0 KB/s
Wrote blk_block=10000, dev_blk_num=9999 VolBytes=645,055,424 rate=2744.9 KB/s
Wrote blk_block=15000, dev_blk_num=14999 VolBytes=967,615,368 rate=2748.9 KB/s
Wrote blk_block=20000, dev_blk_num=4499 VolBytes=1,290,175,312 rate=2733.4
KB/s
Wrote blk_block=25000, dev_blk_num=9499 VolBytes=1,612,735,264 rate=2733.4
KB/s
Wrote blk_block=30000, dev_blk_num=14499 VolBytes=1,935,295,208 rate=2737.3
KB/s
16:57:11 Flush block, write EOF
Wrote blk_block=35000, dev_blk_num=3000 VolBytes=2,257,855,160 rate=2720.3
KB/s
Wrote blk_block=40000, dev_blk_num=8000 VolBytes=2,580,415,104 rate=2722.0
KB/s
Wrote blk_block=45000, dev_blk_num=13000 VolBytes=2,902,975,048 rate=2725.8
KB/s
Wrote blk_block=50000, dev_blk_num=2500 VolBytes=3,225,535,000 rate=2722.0
KB/s
Wrote blk_block=55000, dev_blk_num=7500 VolBytes=3,548,094,944 rate=2723.0
KB/s
Wrote blk_block=60000, dev_blk_num=12500 VolBytes=3,870,654,888 rate=2725.8
KB/s
17:09:51 Flush block, write EOF
Wrote blk_block=65000, dev_blk_num=1000 VolBytes=4,193,214,840 rate=2717.6
KB/s
Wrote blk_block=70000, dev_blk_num=6000 VolBytes=4,515,774,784 rate=2718.7
KB/s
Wrote blk_block=75000, dev_blk_num=11000 VolBytes=4,838,334,736 rate=2721.2
KB/s
Wrote blk_block=80000, dev_blk_num=500 VolBytes=5,160,894,680 rate=2719.1 KB/s
Wrote blk_block=85000, dev_blk_num=5500 VolBytes=5,483,454,624 rate=2720.0
KB/s
Wrote blk_block=90000, dev_blk_num=10500 VolBytes=5,806,014,576 rate=2722.0
KB/s
Wrote blk_block=95000, dev_blk_num=15500 VolBytes=6,128,574,520 rate=2723.8
KB/s
17:22:31 Flush block, write EOF
Wrote blk_block=100000, dev_blk_num=4000 VolBytes=6,451,134,472 rate=2717.4
KB/s
Wrote blk_block=105000, dev_blk_num=9000 VolBytes=6,773,694,416 rate=2719.3
KB/s
Wrote blk_block=110000, dev_blk_num=14000 VolBytes=7,096,254,360 rate=2721.0
KB/s
Wrote blk_block=115000, dev_blk_num=3500 VolBytes=7,418,814,312 rate=2718.5
KB/s
Wrote blk_block=120000, dev_blk_num=8500 VolBytes=7,741,374,256 rate=2720.1 KB
************************************************************
I get 44 MB/sec average in real use. Below I describe my configuration,
which (hopefully) is somewhat helpful.
My dedicated bacula server is a dual Athlon MP 2GzH system with 4 SATA
drives (add-in non-RAID 4 port Promise controller) and a Gig NIC running
Linux Ubuntu 5.10. I have the same version of bacula, and it is also
compiled from source.
The main performance enhancement for me was switching to a system with
more CPU power. The second most important performance enhancement was
how I configured the systems four drives, so as to not be disk I/O bound
(although that's still the slows part of the system). Two of the drives
are configured with software RAID1 for the OS etc. The other two (250GB
each) are configured as software RAID0 with XFS, and serve as a
dedicated data spool for bacula. I also spool attributes so that data
entry into the catalog does not slow writing to the tape.
With this configuration I'm getting between 40 and 54 MB/sec write to
the tape drive during the de-spooling process with an average around 44
MB/sec. During that time bacula uses about 50% of the system's CPU
power (I think this may be because I have md5 sums turned on, but I'm
not sure). I believe the advertised speed for LTO3 is around 60 MB/sec,
but I think seeing a 44 MB/sec average in real world performance (on a
relatively old and cheap server) is pretty decent.
Also, I'm using an older SCSI controller card (160 not 320) and have
never tried to tune the drive with mt or any other utility.
Keith
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
