From de6c806af7e289c11196a53c0e1ff1736f3c2e8e Mon Sep 17 00:00:00 2001
From: Michele Cereda <cereda.michele@gmail.com>
Date: Sun, 12 Oct 2025 07:48:51 +0200
Subject: [PATCH] chore(dblab): destroy automatic snapshot, manually

---
 knowledge base/dblab engine.md | 43 ++++++++++++++++++++++++++++++++++
 snippets/dblab.fish            |  4 ++++
 2 files changed, 47 insertions(+)

diff --git a/knowledge base/dblab engine.md b/knowledge base/dblab engine.md
index da9fb8f..2cf16a8 100644
--- a/knowledge base/dblab engine.md	
+++ b/knowledge base/dblab engine.md	
@@ -14,6 +14,7 @@ The [website] hosts the SaaS version.
    1. [Clean up](#clean-up)
 1. [Automatically full refresh data without downtime](#automatically-full-refresh-data-without-downtime)
 1. [Troubleshooting](#troubleshooting)
+   1. [Cannot destroy automatic snapshot in the pool](#cannot-destroy-automatic-snapshot-in-the-pool)
    1. [The automatic full refresh fails claiming it cannot find available pools](#the-automatic-full-refresh-fails-claiming-it-cannot-find-available-pools)
 1. [Further readings](#further-readings)
    1. [Sources](#sources)
@@ -395,6 +396,48 @@ without downtime.
 
 ## Troubleshooting
 
+### Cannot destroy automatic snapshot in the pool
+
+Error message example:
+
+```plaintext
+2025/10/07 09:49:32 cannot destroy automatic snapshot in the pool
+```
+
+Root cause: still unknown.
+
+Short term solution: manually delete the ZFS snapshots and restart the Engine.
+
+<details>
+
+1. Decide what snapshots need to be deleted.
+
+   ```sh
+   $ zfs list -t 'snapshot'
+   NAME                                   USED  AVAIL  REFER  MOUNTPOINT
+   dblab_pool_0@snapshot_20250924055533  92.3G      -   266G  -
+   dblab_pool_1@snapshot_20250923130042   132G      -   144G  -
+   dblab_pool_1@snapshot_20250915224319   142G      -   145G  -
+   dblab_pool_1@snapshot_20251002175419  87.5K      -   145G  -
+   ```
+
+1. Ensure no clone is using those snapshots.\
+   Reset those that do if necessary.
+1. Destroy the chosen ZFS snapshots.
+
+   ```sh
+   sudo zfs destroy 'dblab_pool_1@snapshot_20250923130042'
+   ```
+
+1. Restart the DBLab Engine's container.\
+   Needed to make it recognize the snapshots are gone.
+
+   ```sh
+   sudo docker container restart 'dblab_server'
+   ```
+
+</details>
+
 ### The automatic full refresh fails claiming it cannot find available pools
 
 Root cause: in version 4.0.0, the DBLab Engine happened to consider a pool used by clones, even if those clones were
diff --git a/snippets/dblab.fish b/snippets/dblab.fish
index b75589b..dfeda3a 100644
--- a/snippets/dblab.fish
+++ b/snippets/dblab.fish
@@ -122,6 +122,10 @@ dblab clone destroy 'some-clone'
 curl -X 'DELETE' 'https://dblab.example.org:1234/api/clone/some-clone' \
 	-H "Verification-Token: $(gopass show -o 'dblab')"
 
+# Delete snapshots
+# Cannot currently delete automatic snapshots via API
+zfs destroy 'dblab_pool_1@snapshot_20250826190434'
+
 # Get admin config in YAML format
 curl 'https://dblab.example.org:1234/api/admin/config.yaml' -H "Verification-Token: $(gopass show -o 'dblab')"