shaba/lvm2
1
0
Fork 0
mirror of git://sourceware.org/git/lvm2.git synced 2025-01-03 05:18:29 +03:00
Code Releases Activity

tests: reduce sizes in pvck-dump and improve checks

Browse Source 
Smaller devs can be used so tests can be run on small vms.
Improve checks.
This commit is contained in:
David Teigland 2020-03-04 11:30:50 -06:00
parent f6667f94cb
commit 4b5bfa779a
1 changed files with 50 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

66
test/shell/pvck-dump.sh
View File

@ -12,13 +12,16 @@
. lib/inittest
aux prepare_devs 8 512
# 4 devs each 128MB
aux prepare_devs 4 128
get_devs
dd if=/dev/zero of="$dev1" || true
dd if=/dev/zero of="$dev2" || true
dd if=/dev/zero of="$dev1" bs=1M count=32 || true
dd if=/dev/zero of="$dev2" bs=1M count=32 || true
dd if=/dev/zero of="$dev3" bs=1M count=32 || true
dd if=/dev/zero of="$dev4" bs=1M count=32 || true
# clear entire dev to cover mda2
dd if=/dev/zero of="$dev3" || true
dd if=/dev/zero of="$dev4" || true
pvcreate "$dev1"
pvcreate "$dev2"
@ -27,6 +30,8 @@ pvcreate --pvmetadatacopies 0 "$dev4"
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
pvs
pvck --dump headers "$dev1" > h1
pvck --dump headers "$dev2" > h2
pvck --dump headers "$dev3" > h3
@ -129,14 +134,17 @@ diff area1 area3b
vgremove -ff $vg
pvremove "$dev1"
pvremove "$dev2"
pvremove "$dev3"
pvremove "$dev4"
dd if=/dev/zero of="$dev1" bs=1M count=32 || true
dd if=/dev/zero of="$dev2" bs=1M count=32 || true
dd if=/dev/zero of="$dev3" bs=1M count=32 || true
dd if=/dev/zero of="$dev4" bs=1M count=32 || true
# clear entire dev to cover mda2
dd if=/dev/zero of="$dev1" || true
pvcreate --pvmetadatacopies 2 --metadatasize 32M "$dev1"
vgcreate $SHARED -s 512K --metadatasize 32M $vg "$dev1" "$dev2" "$dev3" "$dev4" "$dev5" "$dev6" "$dev7" "$dev8"
vgcreate $SHARED -s 512K --metadatasize 32M $vg "$dev1" "$dev2" "$dev3" "$dev4"
for i in `seq 1 500`; do lvcreate -an -n lv$i -l1 $vg; done
@ -145,20 +153,46 @@ pvck --dump headers "$dev1" > h1
pvck --dump metadata_search "$dev1" > m1
grep "seqno 500" m1
dd if="$dev1" of=dev1dd bs=1M count=32
# When metadatasize is 32M, headers/rounding can mean that
# we need more than the first 32M of the dev to get all the
# metadata.
dd if="$dev1" of=dev1dd bs=1M count=34
# Clear the header so that we force metadata_search to use
# the settings instead of getting the mda_size/mda_offset
# from the headers.
dd if=/dev/zero of="$dev1" bs=4K count=1
# mda_size for mda1 is 32M - 4K
pvck --dump metadata_search --settings "mda_num=1 mda_size=33550336" "$dev1" > m1b
# Warning: these checks are based on copying specific numbers
# seen when running these commands, but these numbers could
# change as side effects of other things. That makes this
# somewhat fragile, and we might want to remove some of the
# these checks if they are hard to keep working.
# by experimentation, mda_size for mda1 is 34598912
pvck --dump metadata_search --settings "mda_num=1 mda_size=34598912" "$dev1" > m1b
# by experimentation, metadata 484 is the last in the mda1 buffer
grep "seqno 484" m1b
# by experimentation, metadata 485 is the last in the mda1 buffer
grep "seqno 485" m1b
grep "seqno 500" m1b
pvck --dump metadata_search --settings "mda_num=1 mda_size=33550336" dev1dd > m1c
grep "seqno 500" m1c
# same results when using file as on device
pvck --dump metadata_search --settings "mda_num=1 mda_size=34598912" dev1dd > m1c
# by experimentation, metadata 484 is the last in the mda1 buffer
grep "seqno 484" m1b
# by experimentation, metadata 485 is the last in the mda1 buffer
grep "seqno 485" m1b
grep "seqno 500" m1b
# mda_size for mda2 is 32M
# by experimentation, mda_size for mda2 is 33554432
pvck --dump metadata_search --settings "mda_num=2 mda_size=33554432" "$dev1" > m2
# by experimentation, metadata 477 is the last in the mda2 buffer
grep "seqno 477" m1b
# by experimentation, metadata 478 is the last in the mda2 buffer
grep "seqno 478" m1b
grep "seqno 500" m2
dd if=dev1dd of="$dev1" bs=1M count=32
dd if=dev1dd of="$dev1" bs=4K count=1
vgremove -ff $vg