rust/neon
1
0
Fork 0
mirror of https://github.com/neondatabase/neon.git synced 2026-10-30 07:17:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

storage controller: API + CLI for migrating secondary locations (#10284)

Browse Source 
## Problem

Currently, if we want to move a secondary there isn't a neat way to do
that: we just have migration API for the attached location, and it is
only clean to use that if you've manually created a secondary via
pageserver API in the place you're going to move it to.

Secondary migration API enables:
- Moving the secondary somewhere because we would like to later move the
attached location there.
- Move the secondary location because we just want to reclaim some disk
space from its current location.

## Summary of changes

- Add `/migrate_secondary` API
- Add `tenant-shard-migrate-secondary` CLI
- Add tests for above
This commit is contained in:
John Spray
2025-01-13 14:52:43 +00:00
committed by GitHub
parent ceacc29609
commit ef8bfacd6b
7 changed files with 210 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files
test_runner/regress/test_storage_controller.py
+80 -16
View File
@@ -1052,7 +1052,7 @@ def test_storage_controller_debug_apis(neon_env_builder: NeonEnvBuilder):
that just hits the endpoints to check that they don't bitrot.
"""
neon_env_builder.num_pageservers = 2
neon_env_builder.num_pageservers = 3
env = neon_env_builder.init_start()
tenant_id = TenantId.generate()
@@ -1077,7 +1077,7 @@ def test_storage_controller_debug_apis(neon_env_builder: NeonEnvBuilder):
"GET", f"{env.storage_controller_api}/debug/v1/scheduler"
)
# Two nodes, in a dict of node_id->node
assert len(response.json()["nodes"]) == 2
assert len(response.json()["nodes"]) == 3
assert sum(v["shard_count"] for v in response.json()["nodes"].values()) == 3
assert all(v["may_schedule"] for v in response.json()["nodes"].values())
@@ -1088,13 +1088,25 @@ def test_storage_controller_debug_apis(neon_env_builder: NeonEnvBuilder):
headers=env.storage_controller.headers(TokenScope.ADMIN),
)
# Secondary migration API: superficial check that it migrates
secondary_dest = env.pageservers[2].id
env.storage_controller.request(
"PUT",
f"{env.storage_controller_api}/control/v1/tenant/{tenant_id}-0002/migrate_secondary",
headers=env.storage_controller.headers(TokenScope.ADMIN),
json={"tenant_shard_id": f"{tenant_id}-0002", "node_id": secondary_dest},
)
assert env.storage_controller.tenant_describe(tenant_id)["shards"][0]["node_secondary"] == [
secondary_dest
]
# Node unclean drop API
response = env.storage_controller.request(
"POST",
f"{env.storage_controller_api}/debug/v1/node/{env.pageservers[1].id}/drop",
headers=env.storage_controller.headers(TokenScope.ADMIN),
)
assert len(env.storage_controller.node_list()) == 1
assert len(env.storage_controller.node_list()) == 2
# Tenant unclean drop API
response = env.storage_controller.request(
@@ -1812,7 +1824,13 @@ def test_storcon_cli(neon_env_builder: NeonEnvBuilder):
"""
output_dir = neon_env_builder.test_output_dir
shard_count = 4
env = neon_env_builder.init_start(initial_tenant_shard_count=shard_count)
neon_env_builder.num_pageservers = 2
env = neon_env_builder.init_configs()
env.start()
tenant_id = TenantId.generate()
env.create_tenant(tenant_id, placement_policy='{"Attached":1}', shard_count=shard_count)
base_args = [env.neon_binpath / "storcon_cli", "--api", env.storage_controller_api]
def storcon_cli(args):
@@ -1841,7 +1859,7 @@ def test_storcon_cli(neon_env_builder: NeonEnvBuilder):
# List nodes
node_lines = storcon_cli(["nodes"])
# Table header, footer, and one line of data
assert len(node_lines) == 5
assert len(node_lines) == 7
assert "localhost" in node_lines[3]
# Pause scheduling onto a node
@@ -1859,10 +1877,21 @@ def test_storcon_cli(neon_env_builder: NeonEnvBuilder):
storcon_cli(["node-configure", "--node-id", "1", "--availability", "offline"])
assert "Offline" in storcon_cli(["nodes"])[3]
# Restore node, verify status changes in CLI output
env.pageservers[0].start()
def is_online():
assert "Offline" not in storcon_cli(["nodes"])
wait_until(is_online)
# Let everything stabilize after node failure to avoid interfering with subsequent steps
env.storage_controller.reconcile_until_idle(timeout_secs=10)
# List tenants
tenant_lines = storcon_cli(["tenants"])
assert len(tenant_lines) == 5
assert str(env.initial_tenant) in tenant_lines[3]
assert str(tenant_id) in tenant_lines[3]
# Setting scheduling policies intentionally result in warnings, they're for rare use.
env.storage_controller.allowed_errors.extend(
@@ -1870,23 +1899,58 @@ def test_storcon_cli(neon_env_builder: NeonEnvBuilder):
)
# Describe a tenant
tenant_lines = storcon_cli(["tenant-describe", "--tenant-id", str(env.initial_tenant)])
tenant_lines = storcon_cli(["tenant-describe", "--tenant-id", str(tenant_id)])
assert len(tenant_lines) >= 3 + shard_count * 2
assert str(env.initial_tenant) in tenant_lines[0]
assert str(tenant_id) in tenant_lines[0]
# Migrate an attached location
def other_ps_id(current_ps_id):
return (
env.pageservers[0].id
if current_ps_id == env.pageservers[1].id
else env.pageservers[1].id
)
storcon_cli(
[
"tenant-shard-migrate",
"--tenant-shard-id",
f"{tenant_id}-0004",
"--node",
str(
other_ps_id(
env.storage_controller.tenant_describe(tenant_id)["shards"][0]["node_attached"]
)
),
]
)
# Migrate a secondary location
storcon_cli(
[
"tenant-shard-migrate-secondary",
"--tenant-shard-id",
f"{tenant_id}-0004",
"--node",
str(
other_ps_id(
env.storage_controller.tenant_describe(tenant_id)["shards"][0][
"node_secondary"
][0]
)
),
]
)
# Pause changes on a tenant
storcon_cli(["tenant-policy", "--tenant-id", str(env.initial_tenant), "--scheduling", "stop"])
storcon_cli(["tenant-policy", "--tenant-id", str(tenant_id), "--scheduling", "stop"])
assert "Stop" in storcon_cli(["tenants"])[3]
# Cancel ongoing reconcile on a tenant
storcon_cli(
["tenant-shard-cancel-reconcile", "--tenant-shard-id", f"{env.initial_tenant}-0104"]
)
storcon_cli(["tenant-shard-cancel-reconcile", "--tenant-shard-id", f"{tenant_id}-0104"])
# Change a tenant's placement
storcon_cli(
["tenant-policy", "--tenant-id", str(env.initial_tenant), "--placement", "secondary"]
)
storcon_cli(["tenant-policy", "--tenant-id", str(tenant_id), "--placement", "secondary"])
assert "Secondary" in storcon_cli(["tenants"])[3]
# Modify a tenant's config
@@ -1894,7 +1958,7 @@ def test_storcon_cli(neon_env_builder: NeonEnvBuilder):
[
"patch-tenant-config",
"--tenant-id",
str(env.initial_tenant),
str(tenant_id),
"--config",
json.dumps({"pitr_interval": "1m"}),
]