Q: allocation of PEs to LVs

How are PEs allocated when LVs are created using lvcreate? I created a LV unsing lvcreate from LVM-1.0.1 on an empty VG and was surprised that PEs are allocated starting with PE 1 and not PE 0 as I expected:

isnogud:root# pvcreate /dev/sdd1
pvcreate -- physical volume "/dev/sdd1" successfully created

isnogud:root# vgcreate vg1 /dev/sdd1
vgcreate -- INFO: using default physical extent size 4.00 MB vgcreate -- INFO: maximum logical volume size is 255.99 Gigabyte vgcreate -- doing automatic backup of volume group "vg1" vgcreate -- volume group "vg1" successfully created and activated

isnogud:root# lvcreate -n test vg1 -L32M
lvcreate -- doing automatic backup of "vg1"
lvcreate -- logical volume "/dev/vg1/test" successfully created

isnogud:root# lvdisplay -v /dev/vg1/test
--- Logical volume ---
LV Name /dev/vg1/test
VG Name vg1
LV Write Access read/write
LV Status available
LV # 1
# open 0
LV Size 32.00 MB
Current LE 8
Allocated LE 8
Allocation next free
Read ahead sectors 10000
Block device 58:11

--- Distribution of logical volume on 1 physical volume ---
PV Name PE on PV reads writes
/dev/sdd1 8 0 4

--- logical volume i/o statistic ---
0 reads 4 writes

--- Logical extents ---
LE PV PE reads writes
00000 /dev/sdd1 00001 0 4
00001 /dev/sdd1 00002 0 0
00002 /dev/sdd1 00003 0 0
00003 /dev/sdd1 00004 0 0
00004 /dev/sdd1 00005 0 0
00005 /dev/sdd1 00006 0 0
00006 /dev/sdd1 00007 0 0
00007 /dev/sdd1 00008 0 0

isnogud:root#

The first LV in my other VG starts at PE 0 as expected. I created that VG and LVs using vgcreate and lvcreate from LVM-0.9.1_beta7:

isnogud:root# lvdisplay -v /dev/vg0/root
--- Logical volume ---
LV Name /dev/vg0/root
VG Name vg0
LV Write Access read/write
LV Status available
LV # 1
# open 1
LV Size 128.00 MB
Current LE 32
Allocated LE 32
Stripes 2
Stripe size (KByte) 4
Allocation next free
Read ahead sectors 120
Block device 58:0

--- Distribution of logical volume on 2 physical volumes ---
PV Name PE on PV reads writes
/dev/sda2 16 41814 100298
/dev/sdb2 16 42184 161941

--- logical volume i/o statistic ---
83998 reads 262239 writes

--- Logical extents ---
LE PV PE reads writes
00000 /dev/sda2 00000 2814 663
00001 /dev/sda2 00001 978 24114
...
00014 /dev/sda2 00014 1955 223
00015 /dev/sda2 00015 4050 277
00016 /dev/sdb2 00000 2835 634
00017 /dev/sdb2 00001 946 72276
...
00030 /dev/sdb2 00014 1984 205
00031 /dev/sdb2 00015 4083 311

isnogud:root#

Can someone explain this, please? Is it intentional?

Also, the "Read ahead sectors" seem to have changed from 120 to 10000, which seems quite a lot to me. 10000 sectors is roughly 5 MB. Does every read really cause the next 5 MB to be read from the disk(s)?

urs

Urs,

that is a display related bug.
Thanks for pointing this out.

Please try the following patch:

diff -u -B -r1.3.2.9 lv_show.c
@@ -327,16 +327,18 @@
             "   LE    PV                        PE     reads      writes\n");
    dev = 0;
    for ( pe = 0; pe < lv->lv_allocated_le; pe++) {
+      int pv_index;
+
       if ( lv->lv_current_pe[pe].dev != dev) {
          pv_name = pv_create_name_from_kdev_t ( lv->lv_current_pe[pe].dev);
          dev = lv->lv_current_pe[pe].dev;
       }
+      pv_index = pv_get_index_by_kdev_t ( vg, lv->lv_current_pe[pe].dev);
       printf ( "   %05d %-25s %05u  %-9u  %-9u\n",
                pe,
                pv_name,
                ( lv->lv_current_pe[pe].pe -
-                 vg->pv[pv_get_index_by_kdev_t
-                        ( vg, lv->lv_current_pe[pe].dev)]->pe_on_disk.base /
+                 LVM_VGDA_SIZE(vg->pv[pv_index]) /
                  SECTOR_SIZE) /
                  ( lv->lv_size / lv->lv_allocated_le),
                  lv->lv_current_pe[pe].reads,

Regards,
Heinz -- The LVM Guy --

*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen Sistina Software Inc.
Senior Consultant/Developer Am Sonnenhang 11
56242 Marienrachdorf Germany
[email protected] +49 2626 141200

                                                       FAX 924446

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Yes, that patch did fix the display bug. Thanx.

urs