[task]
name = "syscraft/vm-ceph-bootstrap"
version = "0.1.0"

[metadata]
author_name = "Yuan Gao"
author_email = "gaoy31@gmail.com"
author_organization = "University of Wisconsin - Madison"

difficulty = "hard"
difficulty_explanation = """\
End-to-end Ceph cluster deployment across 7 KVM VMs: bootstrap with cephadm, \
join 6 worker nodes, bring up OSDs on loop-backed data disks, create both a \
replicated pool (size=3) and an erasure-coded pool (jerasure k=4 m=2 host \
failure domain), then create a 100 GiB RBD image, map it on the primary, \
format XFS, and mount it. A correct deployment cannot be hallucinated — every \
step requires real cluster state to converge.\
"""

solution_explanation = """\
On node0: cephadm --image quay.io/ceph/ceph:v19.2.0 bootstrap --mon-ip <node0 \
private IP from /etc/syscraft/cluster-info> --allow-fqdn-hostname. Distribute \
the cluster SSH key, ceph orch host add node1..node6, ceph orch apply osd \
--all-available-devices (picks up /dev/loop0 and /dev/loop1 on every node), \
ceph osd pool create data_rep then set size=3 and application rbd, define an \
EC profile (plugin=jerasure technique=reed_sol_van k=4 m=2 \
crush-failure-domain=host) and create data_ec from it, rbd create \
data_rep/bench_image --size 102400, rbd map, mkfs.xfs, mount /mnt/ceph_data.\
"""

verification_explanation = """\
Vendored copy of cephbench judge.sh runs on node0 and scores 0–100 across 7 \
checks: cluster reachable (5), HEALTH_OK (15), data_rep replicated size=3 \
app=rbd (15), data_ec jerasure reed_sol_van k=4 m=2 host-domain (15), \
bench_image RBD ~100 GiB (20), /mnt/ceph_data mounted (20), write+read probe \
on the mount (10). Reward = score / 100.\
"""

category = "systems_administration"
tags = [
    "ceph",
    "cephadm",
    "distributed-storage",
    "rbd",
    "erasure-coding",
    "linux",
    "vm-cluster",
]
expert_time_estimate_hours = 1.5

[environment]
type = "vm-cluster"
vm_backend = "libvirt"
base_image = "ubuntu-24.04"
network = "192.168.100.0/24"

[[environment.nodes]]
name = "node0"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-mon", "ceph-mgr", "ceph-osd", "primary"]

[[environment.nodes]]
name = "node1"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[[environment.nodes]]
name = "node2"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[[environment.nodes]]
name = "node3"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[[environment.nodes]]
name = "node4"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[[environment.nodes]]
name = "node5"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[[environment.nodes]]
name = "node6"
cpus = 2
memory_mb = 8192
storage_mb = 102400
roles = ["ceph-osd"]

[agent]
timeout_sec = 1800

[verifier]
timeout_sec = 240