0

I was testing read/write operation on FS to check speed via dd cmd.

/app - created with LVM
/app1 - created with parted

when I run below cmds, can see huge diff in speed for both operation with same cmd. Wanted to know why such discrepancy if both lvm and war partition are alligned similarly. or is it like lvm partition is slower than raw partition?

[root@xxx~]# dd if=/dev/zero of=/app1/testfile bs=1000M count=1 oflag=direct
1+0 records in
1+0 records out
1048576000 bytes (1.0 GB) copied, 5.27156 s, 199 MB/s
[root@xxx~]# dd if=/dev/zero of=/app1/testfile1 bs=1000M count=10 oflag=direct
10+0 records in
10+0 records out
10485760000 bytes (10 GB) copied, 51.5217 s, 204 MB/s
[root@xxx~]# dd if=/dev/zero of=/app1/testfile2 bs=1000M count=20 oflag=direct
20+0 records in
20+0 records out
20971520000 bytes (21 GB) copied, 102.977 s, 204 MB/s
[root@xxx~]# dd if=/dev/sde1 of=/dev/null bs=1G count=1
1+0 records in
1+0 records out
1073741824 bytes (1.1 GB) copied, 2.41581 s, 444 MB/s
[root@xxx~]#
[root@xxx~]# dd if=/dev/sde1 of=/dev/null bs=1G count=10
10+0 records in
10+0 records out
10737418240 bytes (11 GB) copied, 24.4689 s, 439 MB/s

With lvm

[root@xxx~]# dd if=/dev/zero of=/app/testfile1 bs=1000M count=1 oflag=direct
1+0 records in
1+0 records out
1048576000 bytes (1.0 GB) copied, 5.24942 s, 200 MB/s
[root@xxx~]# dd if=/dev/zero of=/app/testfile2 bs=1000M count=10 oflag=direct
10+0 records in
10+0 records out
10485760000 bytes (10 GB) copied, 51.4475 s, 204 MB/s
[root@xxx~]# rm -f /app/testfile2 /app/testfile1
[root@xxx~]# dd if=/dev/mapper/vg00-app of=/dev/null bs=1000M count=1
1+0 records in
1+0 records out
1048576000 bytes (1.0 GB) copied, 4.04793 s, 259 MB/s
[root@xxx~]# dd if=/dev/mapper/vg00-app of=/dev/null bs=1000M count=10
10+0 records in
10+0 records out
10485760000 bytes (10 GB) copied, 43.5869 s, 241 MB/s
[root@xxx ~]#
user640121
  • 1
  • 1
  • Can you format your question so that it becomes readable? Also, please ask a question rather than just making a statement. – berndbausch Apr 20 '21 at 11:40
  • Assuming the question is `Why?` - possibly an artifact of specifying `bs=1000M count=1`. Repeat your test with something in the order of actual blocks of the underlying block device (such as `bs=64k count=16000`). – anx Apr 20 '21 at 12:30
  • 1
    Are you sure your LVM and raw partition are both aligned similarly? Also, I'd prefer to only consider `fio` as performance testing tool, and never use just dd for that. – Nikita Kipriyanov Apr 20 '21 at 12:52

0 Answers0