Add Manual host banning to PgCat (#340)

Sometimes we want an admin to be able to ban a host for some time to route traffic away from that host for reasons like partial outages, replication lag, and scheduled maintenance.

We can achieve this today using a configuration update but a quicker approach is to send a control command to PgCat that bans the replica for some specified duration.

This command does not change the current banning rules like

Primaries cannot be banned
When all replicas are banned, all replicas are unbanned
This commit is contained in:
Mostafa Abdelraouf
2023-03-06 06:10:59 -06:00
committed by GitHub
parent 8a0da10a87
commit 2cc6a09fba
5 changed files with 300 additions and 13 deletions

View File

@@ -287,6 +287,76 @@ describe "Admin" do
end
end
describe "Manual Banning" do
let(:processes) { Helpers::Pgcat.single_shard_setup("sharded_db", 10) }
before do
new_configs = processes.pgcat.current_config
# Prevent immediate unbanning when we ban localhost
new_configs["pools"]["sharded_db"]["shards"]["0"]["servers"][0][0] = "127.0.0.1"
new_configs["pools"]["sharded_db"]["shards"]["0"]["servers"][1][0] = "127.0.0.1"
processes.pgcat.update_config(new_configs)
processes.pgcat.reload_config
end
describe "BAN/UNBAN and SHOW BANS" do
it "bans/unbans hosts" do
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
# Returns a list of the banned addresses
results = admin_conn.async_exec("BAN localhost 10").to_a
expect(results.count).to eq(2)
expect(results.map{ |r| r["host"] }.uniq).to eq(["localhost"])
# Subsequent calls should yield no results
results = admin_conn.async_exec("BAN localhost 10").to_a
expect(results.count).to eq(0)
results = admin_conn.async_exec("SHOW BANS").to_a
expect(results.count).to eq(2)
expect(results.map{ |r| r["host"] }.uniq).to eq(["localhost"])
# Returns a list of the unbanned addresses
results = admin_conn.async_exec("UNBAN localhost").to_a
expect(results.count).to eq(2)
expect(results.map{ |r| r["host"] }.uniq).to eq(["localhost"])
# Subsequent calls should yield no results
results = admin_conn.async_exec("UNBAN localhost").to_a
expect(results.count).to eq(0)
results = admin_conn.async_exec("SHOW BANS").to_a
expect(results.count).to eq(0)
end
it "honors ban duration" do
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
# Returns a list of the banned addresses
results = admin_conn.async_exec("BAN localhost 1").to_a
expect(results.count).to eq(2)
expect(results.map{ |r| r["host"] }.uniq).to eq(["localhost"])
sleep(2)
# After 2 seconds the ban should be lifted
results = admin_conn.async_exec("SHOW BANS").to_a
expect(results.count).to eq(0)
end
it "can handle bad input" do
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
expect { admin_conn.async_exec("BAN").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("BAN a").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("BAN a a").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("BAN a -5").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("BAN a 0").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("BAN a a a").to_a }.to raise_error(PG::SystemError)
expect { admin_conn.async_exec("UNBAN").to_a }.to raise_error(PG::SystemError)
end
end
end
describe "SHOW users" do
it "returns the right users" do
admin_conn = PG::connect(processes.pgcat.admin_connection_string)