OCF_HEARTBEAT_LXC(7)

💖 OCF_HEARTBEAT_LXC — bestie edition 💖

okay so this is ocf_heartbeat_lxc, a resource agent from the OCF resource agents family, and it's the one that lets your cluster babysit LXC containers. yes. containers. in a cluster. we love a crossover episode 🫶

---

✨ NAME ✨

ocf_heartbeat_lxc — she manages LXC containers. that's the whole job description, and she does it flawlessly 💅

---

✨ SYNOPSIS ✨

lxc [start | stop | monitor | meta-data | validate-all]

that's it. five moves. pick one, bestie 🎀

---

✨ DESCRIPTION ✨

So this agent is what lets LXC containers be managed by the cluster. iconic behaviour honestly.

now here's the tea for anyone running lxc versions before 1.0.0 🍵 — back then the container gets stopped with kill -PWR instead of lxc-stop. And the agent is out here 'assuming' that your init system will do a nice orderly shutdown when it gets handed that kill -PWR signal. On a sysvinit setup, that assumption only holds if the container has an inittab file with this line in it:

p0::powerfail:/sbin/init 0

no inittab line, no graceful shutdown. that's not a vibe, that's just a fact 😭

---

✨ SUPPORTED PARAMETERS ✨

these are the knobs, and we are going to turn every single one of them correctly 🔥

• container — the unique name for this 'Container Instance'. like, 'test1' for example. this one is unique, required, string, and has no default — so you will be filling it in, bestie, no skipping 💀
• config — the absolute path to the file that holds the specific configuration for this container. for example '/etc/lxc/test1/config'. also required, string, no default. two required params, we're thriving ✨
• log — the absolute path to the container's log file. this one is optional, string, and defaults to "/run/resource-agents/RESOURCE_ID.log" if you don't say anything 🎀
• use_screen — this gives you the option of capturing the container's 'root console' and showing it on a separate screen. to actually peek at that screen output, run screen -r {container name}. the default is 'false', so if you want this you gotta flip it to 'true' — she will not do it for you 🙈

---

✨ SUPPORTED ACTIONS ✨

this resource agent supports the following actions (operations), and each one has ✨opinions✨ about how long you should let it take:

• start — starts the resource. suggested minimum timeout: 10s.
• stop — stops the resource. suggested minimum timeout: 30s.
• monitor — performs a detailed status check. suggested minimum timeout: 20s, and a suggested interval of 60s.
• validate-all — performs a validation of the resource configuration. suggested minimum timeout: 20s.
• meta-data — retrieves resource agent metadata (internal use only, bestie, don't touch unless you know). suggested minimum timeout: 5s.

---

✨ EXAMPLE CRM SHELL ✨

here's an example config for a lxc resource using the crm(8) shell. copy it, love it, change the strings 🌸

primitive p_lxc ocf:heartbeat:lxc \
  params \
    container=string \
    config=string \
  op monitor timeout="20s" interval="60s" depth="0"

---

✨ EXAMPLE PCS ✨

and here's the same energy via pcs(8) 💜

pcs resource create p_lxc ocf:heartbeat:lxc \
  container=string \
  config=string \
  op monitor timeout="20s" interval="60s" OCF_CHECK_LEVEL="0"

---

✨ SEE ALSO ✨

• http://clusterlabs.org/ — go say hi, they're lovely 🌈

---

✨ AUTHOR ✨

ClusterLabs contributors — and if you want to know about individual authors, go check the resource agent source, the receipts are all there 👑

---

and that's the whole page, no cap. this version of the man page is from resource-agents UNKNOWN, dated 04/08/2026. slay, stay clustered, and remember: monitor at 60s intervals is self care 🎀💖✨🦋