OCF_HEARTBEAT_MEDIAT(7)

💜 ocf_heartbeat_mediationzone(7) — the MediationZone pico resource agent, explained by your favorite cluster bestie 🫶

okay so. bestie. listen. this is the manual page for ocf_heartbeat_mediationzone, and it lives in the OCF resource agents section of the manual. it's giving "niche but powerful" and honestly? we're obsessed. 💅

It's a resource agent — the little helper that teaches Pacemaker how to babysit a MediationZone thing. Let's get into it. ✨

---

🌸 NAME 🌸

ocf_heartbeat_mediationzone — literally just here to manage a MediationZone pico process. that's it, that's the job. 🎀

---

✨ SYNOPSIS ✨

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

those are your moves, bestie. pick one. 🦋

---

💖 DESCRIPTION 💖

so the mediationzone resource agent manages MediationZone pico processes — and MediationZone is by DigitalRoute, for the record — as active/passive cluster resources. iconic behavior honestly. 👑

One resource per pico. one pico, one resource, no sharing, we don't do that here. 🙅‍♀️

The order matters, okay? The platform VIP has to start first, then the platform container, and then any other containers (ui, ec, etc.). it's a whole lineup. think of it as the main character arriving before the entourage. 💫

Now the important part — set <yellow>meta priority=100</yellow> on the platform resource. Why? Because the platform is the most critical pico — literally every other pico depends on it. 😭 So in a split-brain situation, Pacemaker fences the lower-priority node first, which means the platform node survives. slay. protective. 🛡️

⚠️ but bestie, heads up: the cluster property priority-fencing-delay must be greater than 0 for this to actually take effect. if it's 0, none of that protection kicks in. no cap. 🙈

---

🌈 SUPPORTED PARAMETERS 🌈

these are the knobs you get to turn. turn them responsibly. 🎛️

• os_user — the OS user that owns the MediationZone installation and runs the mzsh commands. this is optional, it's a string, and it defaults to "mzadmin". 👤
• mz_home — path to the MediationZone installation directory. it gets passed to mzsh as MZ_HOME, and mzsh is expected to be sitting at MZ_HOME/bin/mzsh. optional, string, default "/opt/mz". 📁
• java_home — path to the JDK or JRE that MediationZone uses. optional, string, default "/usr/lib/jvm/java-17-openjdk". ☕
• pico_name — the name of the MediationZone pico you want to manage. the known names with default ports are platform (port 9000) and ui (port 9001). for any other pico name + port combo, you gotta set pico_port explicitly. optional, string, default "platform". 🏷️
• pico_port — the port the pico web server listens on. it's used to clear TIME-WAIT connections before startup (yes, really, that's a thing 🫠). it defaults to 9000 for platform and 9001 for ui. you must set it explicitly when you're managing a different pico type, or when the default port has been changed in the container configuration. optional, integer, no default. 🔌
• mzsh_timeout — the max time in seconds to wait for a single mzsh call (that's status, shutdown, kill, and the pre-start kill that runs before startup). must be 10 or higher. and here's the tea: the stop action chains shutdown then kill on failure, so worst-case stop time is 2 × mzsh_timeout. if mzsh_timeout is too short, both shutdown and kill time out and the stop action fails — and then Pacemaker fences the node to recover. 😭 also, set the stop operation timeout on the cluster resource to at least 2 × mzsh_timeout + 60s. optional, integer, default 30. ⏱️
• monitor_retries — how many times the monitor action retries mzsh status after an initial failure before reporting the pico as not running or failed. this applies only to the monitor action — start and stop are not affected, okay? a value of 0 disables retries. bump it up if transient status failures are causing unwanted restarts in your environment. must be 0 or higher. optional, integer, default 1. 🔁

---

👑 SUPPORTED ACTIONS 👑

this resource agent supports the following actions (operations), and yes, each one has a suggested minimum timeout so you don't leave it hanging: ⏳

• start — starts the resource. suggested minimum timeout: 360s. 🚀
• stop — stops the resource. suggested minimum timeout: 360s. 🛑
• monitor — performs a detailed status check. suggested minimum timeout: 90s, suggested interval: 60s. 👀
• validate-all — performs a validation of the resource configuration. suggested minimum timeout: 30s. ✅
• meta-data — retrieves resource agent metadata (internal use only). suggested minimum timeout: 5s. 🗂️
• methods — suggested minimum timeout: 5s. 🧪
• usage — suggested minimum timeout: 5s. 📖

---

🧚 EXAMPLE CRM SHELL 🧚

here's an example config for a mediationzone resource using the crm(8) shell. steal it, it's cute:

primitive p_mediationzone ocf:heartbeat:mediationzone \
  op monitor timeout="90s" interval="60s"

---

🍓 EXAMPLE PCS 🍓

and here's the same vibe but for pcs(8):

pcs resource create p_mediationzone ocf:heartbeat:mediationzone \
  op monitor timeout="90s" interval="60s"

---

🦋 SEE ALSO 🦋

• http://clusterlabs.org/ — go look, it's the clusterlabs home base 💙

---

💀 AUTHOR 💀

ClusterLabs contributors — and if you want to know about individual authors, go check the resource agent source, it's all in there. 🫶

---

and that's the page!! 💜 version resource-agents UNKNOWN, dated 10/01/2026 — mwah, sending you cluster-safe kisses, go forth and keep that platform pico alive 🌈✨🎀