|
|
|
|
|
|
|
|
|
|
xen-users
Are you increasing the block size when you copy with dd??? Or are you
just using the default which I believe is 512 bytes. I use dd to create
my lvm's from template files. My template OS's are around 500M to 1G.
For example.
inuyasha:~# time dd if=debian3.1 of=/dev/vds/lvol2 bs=1M
512+0 records in
512+0 records out
536870912 bytes transferred in 25.429652 seconds (21112004 bytes/sec)
real 0m25.432s
user 0m0.000s
sys 0m2.090s
25 seconds to copy a 500 meg file to the lvm and as you can see almost
no cpu time was used. However if I don't increase the blocksize I get this.
inuyasha:~# time dd if=debian3.1 of=/dev/vds/lvol2
1048576+0 records in
1048576+0 records out
536870912 bytes transferred in 153.342399 seconds (3501125 bytes/sec)
real 2m33.345s
user 0m0.650s
sys 0m3.530s
Big difference huh?? 25 seconds to 2.5 minutes. Still doesn't thrash the
processor but this is a dual 2.4ghz xeon with an ultra 160 scsi array.
Dave
Anand wrote:
If i have two volumes test7 and test8. What is most efficient way to
copy data from one to another ?
I did tests using cat, rsync, dd, cpio and all of them seem to be
consuming almost all of my dom0 256MB ram and cpu load goes up to a
max 1.65 (athlon 2.6ghz, 1gb ram)
Infact dd takes more time to copy the data, ofcourse the reason being
the way it operates and copies sector by sector. All other tools
averaged time at 4.5 mins, dd took 6+mins.
The test7 volume is of 850MB and the test8 volume is of 1GB. The test7
volume is actually the template for the centos which is used for
various domains.
Can anyone please share their experiences on what they use to copy
data most efficiently ?
--
regards,
Anand
------------------------------------------------------------------------
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
|
|
|
|