ceph error ERANGE when creating bluestore

[nissekalle]

CentosOS 7.0
Kolla-ansible 7.0.1
binary
rocky
---
1 network node
3 ceph nodes BS
2 computer nodes
2 control nodes
1 monitor node
----
creating 3 x 5 disks (3 OSD:s)....

parted libvirt/images/ceph-bs-01.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_001 1 -1
parted libvirt/images/ceph-bs-block-01.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_001_B 1 -1
parted libvirt/images/ceph-bs-block-wal-01.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_001_W 1 -1
parted libvirt/images/ceph-bs-block-db-01.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_001_D 1 -1
parted libvirt/images/ceph-bs-cache-01.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_CACHE_BOOTSTRAP_BS 1 -1

parted libvirt/images/ceph-bs-02.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_002 1 -1
parted libvirt/images/ceph-bs-block-02.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_002_B 1 -1
parted libvirt/images/ceph-bs-block-wal-02.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_002_W 1 -1
parted libvirt/images/ceph-bs-block-db-02.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_002_D 1 -1
parted libvirt/images/ceph-bs-cache-02.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_CACHE_BOOTSTRAP_BS 1 -1

parted libvirt/images/ceph-bs-03.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_003 1 -1
parted libvirt/images/ceph-bs-block-03.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_003_B 1 -1
parted libvirt/images/ceph-bs-block-wal-03.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_003_W 1 -1
parted libvirt/images/ceph-bs-block-db-03.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP_BS_003_D 1 -1
parted libvirt/images/ceph-bs-cache-03.img -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_CACHE_BOOTSTRAP_BS 1 -1

it seems that the default settings are not enough:

pg_num 128 size 3 would mean 960 total pgs, which exceeds max 750 pgs

should this be change in the defaults?

[nissekalle (nissekalle)]

no problem in stein