Consolidates the logic in the run_check_repl_dbhash.js and run_validate_collections.js for discovering all of the mongod processes in a MongoDB deployment into a new discover_topology.js library. Also adds a test that relies on mongod logging to verify that the run_check_repl_dbhash.js and run_validate_collections.js hooks execute on all of the expected servers.
203 lines
7.7 KiB
JavaScript
203 lines
7.7 KiB
JavaScript
// Runner for checkDBHashes() that runs the dbhash command on all replica set nodes
|
|
// to ensure all nodes have the same data.
|
|
'use strict';
|
|
|
|
(function() {
|
|
load('jstests/libs/discover_topology.js'); // For Topology and DiscoverTopology.
|
|
load('jstests/libs/parallelTester.js'); // For ScopedThread.
|
|
|
|
// A thin wrapper around master/slave nodes that provides methods necessary for checking data
|
|
// consistency between the master and slave nodes.
|
|
//
|
|
// DEPRECATED: This wrapper is only intended to be used for the master-slave deployment started
|
|
// by resmoke.py as part of the master_slave_jscore_passthrough.yml test suite and it shouldn't
|
|
// be used for any other master/slave tests.
|
|
//
|
|
// TODO SERVER-32143: Remove this wrapper.
|
|
function MasterSlaveDBHashTest(primaryHost) {
|
|
const master = new Mongo(primaryHost);
|
|
const masterPort = master.host.split(':')[1];
|
|
const slave = new Mongo('localhost:' + String(parseInt(masterPort) + 1));
|
|
|
|
this.nodeList = function nodeList() {
|
|
return [master.host, slave.host];
|
|
};
|
|
|
|
this.getHashes = function getHashes(db) {
|
|
const combinedRes = {};
|
|
let res = master.getDB(db).runCommand('dbhash');
|
|
assert.commandWorked(res);
|
|
combinedRes.master = res;
|
|
|
|
res = slave.getDB(db).runCommand('dbhash');
|
|
assert.commandWorked(res);
|
|
combinedRes.slaves = [res];
|
|
|
|
return combinedRes;
|
|
};
|
|
|
|
this.getPrimary = function getPrimary() {
|
|
slave.setSlaveOk();
|
|
this.liveNodes = {master: master, slaves: [slave]};
|
|
return master;
|
|
};
|
|
|
|
this.getSecondaries = function getSecondaries() {
|
|
return [slave];
|
|
};
|
|
|
|
this.awaitReplication = function awaitReplication() {
|
|
assert.commandWorked(master.adminCommand({fsyncUnlock: 1}),
|
|
'failed to unlock the primary');
|
|
|
|
print('Starting fsync on master to flush all pending writes');
|
|
assert.commandWorked(master.adminCommand({fsync: 1}));
|
|
print('fsync on master completed');
|
|
|
|
const kTimeout = 5 * 60 * 1000; // 5 minute timeout
|
|
const dbNames = master.getDBNames();
|
|
|
|
for (let dbName of dbNames) {
|
|
if (dbName === 'local') {
|
|
continue;
|
|
}
|
|
|
|
print('Awaiting replication of inserts into ' + dbName);
|
|
assert.writeOK(master.getDB(dbName).await_repl.insert(
|
|
{awaiting: 'repl'}, {writeConcern: {w: 2, wtimeout: kTimeout}}),
|
|
'Awaiting replication failed');
|
|
}
|
|
print('Finished awaiting replication');
|
|
assert.commandWorked(master.adminCommand({fsync: 1, lock: 1}),
|
|
'failed to re-lock the primary');
|
|
};
|
|
|
|
this.checkReplicatedDataHashes = function checkReplicatedDataHashes() {
|
|
const msgPrefix = 'checkReplicatedDataHashes for master-slave deployment';
|
|
const excludedDBs = jsTest.options().excludedDBsFromDBHash || [];
|
|
|
|
// Since UUIDs aren't explicitly replicated in master-slave deployments, we ignore the
|
|
// UUID in the output of the 'listCollections' command to avoid reporting a known data
|
|
// inconsistency issue from checkReplicatedDataHashes().
|
|
const ignoreUUIDs = true;
|
|
|
|
new ReplSetTest({
|
|
nodes: 0
|
|
}).checkReplicatedDataHashes.call(this, msgPrefix, excludedDBs, ignoreUUIDs);
|
|
};
|
|
|
|
this.checkReplicaSet = function checkReplicaSet() {
|
|
new ReplSetTest({nodes: 0}).checkReplicaSet.apply(this, arguments);
|
|
};
|
|
|
|
this.dumpOplog = function dumpOplog() {
|
|
print('Not dumping oplog for master-slave deployment');
|
|
};
|
|
}
|
|
|
|
function isMasterSlaveDeployment(conn) {
|
|
const cmdLineOpts = assert.commandWorked(conn.adminCommand({getCmdLineOpts: 1}));
|
|
return cmdLineOpts.parsed.master === true;
|
|
}
|
|
|
|
function checkReplicatedDataHashesThread(hosts, testData) {
|
|
try {
|
|
TestData = testData;
|
|
new ReplSetTest(hosts[0]).checkReplicatedDataHashes();
|
|
return {ok: 1};
|
|
} catch (e) {
|
|
return {ok: 0, hosts: hosts, error: e.toString(), stack: e.stack};
|
|
}
|
|
}
|
|
|
|
const startTime = Date.now();
|
|
assert.neq(typeof db, 'undefined', 'No `db` object, is the shell connected to a mongod?');
|
|
|
|
let skipped = false;
|
|
try {
|
|
const conn = db.getMongo();
|
|
|
|
if (isMasterSlaveDeployment(conn)) {
|
|
new MasterSlaveDBHashTest(conn.host).checkReplicatedDataHashes();
|
|
return;
|
|
}
|
|
|
|
const topology = DiscoverTopology.findConnectedNodes(conn);
|
|
|
|
if (topology.type === Topology.kStandalone) {
|
|
print('Skipping data consistency checks for cluster because we are connected to a' +
|
|
' stand-alone mongod: ' + tojsononeline(topology));
|
|
skipped = true;
|
|
return;
|
|
}
|
|
|
|
if (topology.type === Topology.kReplicaSet) {
|
|
if (topology.nodes.length === 1) {
|
|
print('Skipping data consistency checks for cluster because we are connected to a' +
|
|
' 1-node replica set: ' + tojsononeline(topology));
|
|
skipped = true;
|
|
return;
|
|
}
|
|
|
|
new ReplSetTest(topology.nodes[0]).checkReplicatedDataHashes();
|
|
return;
|
|
}
|
|
|
|
if (topology.type !== Topology.kShardedCluster) {
|
|
throw new Error('Unrecognized topology format: ' + tojson(topology));
|
|
}
|
|
|
|
const threads = [];
|
|
try {
|
|
if (topology.configsvr.nodes.length > 1) {
|
|
const thread = new ScopedThread(
|
|
checkReplicatedDataHashesThread, topology.configsvr.nodes, TestData);
|
|
threads.push(thread);
|
|
thread.start();
|
|
} else {
|
|
print('Skipping data consistency checks for 1-node CSRS: ' +
|
|
tojsononeline(topology));
|
|
}
|
|
|
|
for (let shardName of Object.keys(topology.shards)) {
|
|
const shard = topology.shards[shardName];
|
|
|
|
if (shard.type === Topology.kStandalone) {
|
|
print('Skipping data consistency checks for stand-alone shard: ' +
|
|
tojsononeline(topology));
|
|
continue;
|
|
}
|
|
|
|
if (shard.type !== Topology.kReplicaSet) {
|
|
throw new Error('Unrecognized topology format: ' + tojson(topology));
|
|
}
|
|
|
|
if (shard.nodes.length > 1) {
|
|
const thread =
|
|
new ScopedThread(checkReplicatedDataHashesThread, shard.nodes, TestData);
|
|
threads.push(thread);
|
|
thread.start();
|
|
} else {
|
|
print('Skipping data consistency checks for 1-node replica set shard: ' +
|
|
tojsononeline(topology));
|
|
}
|
|
}
|
|
} finally {
|
|
// Wait for each thread to finish. Throw an error if any thread fails.
|
|
const returnData = threads.map(thread => {
|
|
thread.join();
|
|
return thread.returnData();
|
|
});
|
|
|
|
returnData.forEach(res => {
|
|
assert.commandWorked(res, 'data consistency checks failed');
|
|
});
|
|
}
|
|
} finally {
|
|
if (!skipped) {
|
|
const totalTime = Date.now() - startTime;
|
|
print('Finished data consistency checks for cluster in ' + totalTime + ' ms.');
|
|
}
|
|
}
|
|
})();
|