[CentOS] anaconda, kickstart, lvm over raid, logvol --grow, centos7 mystery
Maxim Shpakov
maxim at osetia.org
Thu Jul 31 09:01:51 UTC 2014
Hi!
I can confirm this.
--grow on LVM partition is broken for raid+lvm kickstart installs.
bootloader --location=mbr --driveorder=sda,sdb --append="net.ifnames=0
crashkernel=auto rhgb quiet"
zerombr
clearpart --all --drives=sda,sdb --initlabel
part raid.1 --asprimary --size=200 --ondisk=sda
part raid.2 --size=1 --grow --ondisk=sda
part raid.3 --asprimary --size=200 --ondisk=sdb
part raid.4 --size=1 --grow --ondisk=sdb
raid /boot --fstype=ext4 --level=RAID1 --device=md0 raid.1 raid.3
raid pv.1 --level=RAID1 --device=md1 raid.2 raid.4
volgroup vg0 --pesize=65536 pv.1
logvol swap --name=swap --vgname=vg0 --size=4096
logvol /tmp --fstype=ext4 --name=tmp --vgname=vg0 --size=4096
--fsoptions="noexec,nosuid,nodev,noatime"
logvol / --fstype=ext4 --name=root --vgname=vg0 --size=10240 --grow
--fsoptions="defaults,noatime"
Such partitioning scheme is now working. Anaconda is complaining about
"ValueError: not enough free space in volume group"
Buf if I remove --grow from last logvol - everything is ok.
I don't understand what I'm doing wrong, such kickstart works
flawlessly for C6 installs.
2014-07-16 14:21 GMT+03:00 Borislav Andric <borislav.andric at gmail.com>:
> I am testing some kickstarts on ESXi virtual machine with pair of 16GB disks.
> Partitioning is lvm over raid.
>
> If i am using "logvol --grow i get "ValueError: not enough free space in volume group"
> Only workaround i can find is to add --maxsize=XXX where XXX is at least 640MB less than available.
> (10 extents or 320Mb per created logical volume)
>
> Following snippet is failing with "DEBUG blivet: failed to set size: 640MB short"
>
> part raid.01 --size 512 --asprimary --ondrive=sda
> part raid.02 --size 1 --asprimary --ondrive=sda --grow
> part raid.11 --size 512 --asprimary --ondrive=sdb
> part raid.12 --size 1 --asprimary --ondrive=sdb --grow
> raid /boot --fstype="xfs" --device="md0" --level=RAID1 raid.01 raid.11
> raid pv.01 --fstype="lvmpv" --device="md1" --level=RAID1 raid.02 raid.12
> volgroup vg0 pv.01
> logvol / --fstype="xfs" --grow --size=4096 --name=lvRoot --vgname=vg0
> logvol swap --fstype="swap" --size=2048 --name=lvSwap --vgname=vg0
>
> If i only add --maxsize=13164 everything is working.
> (but after install i have 640MB in 20 Free PE in vg0, for details see "after --maxsize install")
>
> logvol / --fstype="xfs" --grow --size=4096 --name=lvRoot --vgname=vg0
> ------changed to ----->
> logvol / --fstype="xfs" --grow --size=4096 --name=lvRoot --vgname=vg0 --maxsize=13164
>
>
> Some interesting DEBUG lines :
>
>> 15840MB lvmvg vg0 (26)
>> vg0 size is 15840MB
>> Adding vg0-lvRoot/4096MB to vg0
>> vg vg0 has 11424MB free
>
> should it be 11744 or there is 320MB overhead ?
>
>> Adding vg0-lvSwap/2048MB to vg0
>> vg vg0 has 9056MB free
>
> 320MB missing again, total of 640MB
>
>> vg vg0: 9056MB free ; lvs: ['lvRoot', 'lvSwap']
>
> nice, i have 9056MB free in vg0 (640MB short but still ... )
>
>> 1 requests and 303 (9696MB) left in chunk
>> adding 303 (9696MB) to 27 (vg0-lvRoot)
>
> wtf, who is counting what !!
>
>> failed to set size: 640MB short
>
>
> Could anyone shed some light ?
>
>
>
>
>
> P.S.
>
> "after --maxsize install"
> =========================
> If i limit root logvol with --maxsize=13164, after installation i get 640MB of free space (20 Free PE).
>
>
> Missing 640Mb is free according to lvm :
> [root at c7-pxe-install ~]# pvdisplay
> --- Physical volume ---
> PV Name /dev/md1
> VG Name vg0
> PV Size 15.49 GiB / not usable 22.88 MiB
> Allocatable yes
> PE Size 32.00 MiB
> Total PE 495
>>>>>>>>>>>Free PE 20<<<<<<<<<<<<
> Allocated PE 475
> PV UUID uBLBqQ-Tpao-yPVj-1FVA-488x-Bs0K-ebQOmI
>
>
> And i can use it :
> [root at c7-pxe-install ~]# lvextend -L +640M vg0/lvRoot
> Extending logical volume lvRoot to 13.47 GiB
> Logical volume lvRoot successfully resized
>
> [root at c7-pxe-install ~]# xfs_growfs /
> meta-data=/dev/mapper/vg0-lvRoot isize=256 agcount=4, agsize=841728 blks
> = sectsz=512 attr=2, projid32bit=1
> = crc=0
> data = bsize=4096 blocks=3366912, imaxpct=25
> = sunit=0 swidth=0 blks
> naming =version 2 bsize=4096 ascii-ci=0 ftype=0
> log =internal bsize=4096 blocks=2560, version=2
> = sectsz=512 sunit=0 blks, lazy-count=1
> realtime =none extsz=4096 blocks=0, rtextents=0
> data blocks changed from 3366912 to 3530752
> [root at c7-pxe-install ~]# pvdisplay
> --- Physical volume ---
> PV Name /dev/md1
> VG Name vg0
> PV Size 15.49 GiB / not usable 22.88 MiB
> Allocatable yes (but full)
> PE Size 32.00 MiB
> Total PE 495
> Free PE 0
> Allocated PE 495
> PV UUID uBLBqQ-Tpao-yPVj-1FVA-488x-Bs0K-ebQOmI
>
> [root at c7-pxe-install ~]#
> _______________________________________________
> CentOS mailing list
> CentOS at centos.org
> http://lists.centos.org/mailman/listinfo/centos
More information about the CentOS
mailing list