dead_pool_reaper_test.go

package work

import (
	"testing"
	"time"

	"github.com/gomodule/redigo/redis"
	"github.com/stretchr/testify/assert"
)

func TestDeadPoolReaper(t *testing.T) {
	pool := newTestPool(t)
	ns := "work"
	cleanKeyspace(ns, pool)

	conn := pool.Get()
	defer conn.Close()

	workerPoolsKey := redisKeyWorkerPools(ns)

	// Create redis data
	var err error
	err = conn.Send("SADD", workerPoolsKey, "1")
	assert.NoError(t, err)
	err = conn.Send("SADD", workerPoolsKey, "2")
	assert.NoError(t, err)
	err = conn.Send("SADD", workerPoolsKey, "3")
	assert.NoError(t, err)

	err = conn.Send("HMSET", redisKeyHeartbeat(ns, "1"),
		"heartbeat_at", time.Now().Unix(),
		"job_names", "type1,type2",
	)
	assert.NoError(t, err)

	err = conn.Send("HMSET", redisKeyHeartbeat(ns, "2"),
		"heartbeat_at", time.Now().Add(-1*time.Hour).Unix(),
		"job_names", "type1,type2",
	)
	assert.NoError(t, err)

	err = conn.Send("HMSET", redisKeyHeartbeat(ns, "3"),
		"heartbeat_at", time.Now().Add(-1*time.Hour).Unix(),
		"job_names", "type1,type2",
	)
	assert.NoError(t, err)
	err = conn.Flush()
	assert.NoError(t, err)

	// Without this, the assertion for deadPools fails on GitHub CI. Could have
	// something to do with in-memory redis instance.
	time.Sleep(10 * time.Millisecond)

	// Test getting dead pool
	reaper := newDeadPoolReaper(ns, pool, []string{})
	deadPools, err := reaper.findDeadPools()
	assert.NoError(t, err)
	assert.Equal(t, map[string][]string{"2": {"type1", "type2"}, "3": {"type1", "type2"}}, deadPools)

	// Test requeueing jobs
	_, err = conn.Do("lpush", redisKeyJobsInProgress(ns, "2", "type1"), "foo")
	assert.NoError(t, err)
	_, err = conn.Do("incr", redisKeyJobsLock(ns, "type1"))
	assert.NoError(t, err)
	_, err = conn.Do("hincrby", redisKeyJobsLockInfo(ns, "type1"), "2", 1) // worker pool 2 has lock
	assert.NoError(t, err)

	// Ensure 0 jobs in jobs queue
	jobsCount, err := redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Ensure 1 job in inprogress queue
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Reap
	err = reaper.reap()
	assert.NoError(t, err)

	// Ensure 1 jobs in jobs queue
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Ensure 0 job in inprogress queue
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Locks should get cleaned up
	assert.EqualValues(t, 0, getInt64(pool, redisKeyJobsLock(ns, "type1")))
	v, _ := conn.Do("HGET", redisKeyJobsLockInfo(ns, "type1"), "2")
	assert.Nil(t, v)
}

func TestDeadPoolReaperNoHeartbeat(t *testing.T) {
	pool := newTestPool(t)
	ns := "work"

	conn := pool.Get()
	defer conn.Close()

	workerPoolsKey := redisKeyWorkerPools(ns)

	// Create redis data
	var err error
	cleanKeyspace(ns, pool)
	err = conn.Send("SADD", workerPoolsKey, "1")
	assert.NoError(t, err)
	err = conn.Send("SADD", workerPoolsKey, "2")
	assert.NoError(t, err)
	err = conn.Send("SADD", workerPoolsKey, "3")
	assert.NoError(t, err)
	// stale lock info
	err = conn.Send("SET", redisKeyJobsLock(ns, "type1"), 3)
	assert.NoError(t, err)
	err = conn.Send("HSET", redisKeyJobsLockInfo(ns, "type1"), "1", 1)
	assert.NoError(t, err)
	err = conn.Send("HSET", redisKeyJobsLockInfo(ns, "type1"), "2", 1)
	assert.NoError(t, err)
	err = conn.Send("HSET", redisKeyJobsLockInfo(ns, "type1"), "3", 1)
	assert.NoError(t, err)
	err = conn.Flush()
	assert.NoError(t, err)

	// make sure test data was created
	numPools, err := redis.Int(conn.Do("scard", workerPoolsKey))
	assert.NoError(t, err)
	assert.EqualValues(t, 3, numPools)

	// Test getting dead pool ids
	reaper := newDeadPoolReaper(ns, pool, []string{"type1"})
	deadPools, err := reaper.findDeadPools()
	assert.NoError(t, err)
	assert.Equal(t, map[string][]string{"1": {}, "2": {}, "3": {}}, deadPools)

	// Test requeueing jobs
	_, err = conn.Do("lpush", redisKeyJobsInProgress(ns, "2", "type1"), "foo")
	assert.NoError(t, err)

	// Ensure 0 jobs in jobs queue
	jobsCount, err := redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Ensure 1 job in inprogress queue
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Ensure dead worker pools still in the set
	jobsCount, err = redis.Int(conn.Do("scard", redisKeyWorkerPools(ns)))
	assert.NoError(t, err)
	assert.Equal(t, 3, jobsCount)

	// Reap
	err = reaper.reap()
	assert.NoError(t, err)

	// Ensure jobs queue was not altered
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Ensure inprogress queue was not altered
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Ensure dead worker pools were removed from the set
	jobsCount, err = redis.Int(conn.Do("scard", redisKeyWorkerPools(ns)))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Stale lock info was cleaned up using reap.curJobTypes
	assert.EqualValues(t, 0, getInt64(pool, redisKeyJobsLock(ns, "type1")))
	for _, poolID := range []string{"1", "2", "3"} {
		v, _ := conn.Do("HGET", redisKeyJobsLockInfo(ns, "type1"), poolID)
		assert.Nil(t, v)
	}
}

func TestDeadPoolReaperNoJobTypes(t *testing.T) {
	pool := newTestPool(t)
	ns := "work"
	cleanKeyspace(ns, pool)

	conn := pool.Get()
	defer conn.Close()

	workerPoolsKey := redisKeyWorkerPools(ns)

	// Create redis data
	var err error
	err = conn.Send("SADD", workerPoolsKey, "1")
	assert.NoError(t, err)
	err = conn.Send("SADD", workerPoolsKey, "2")
	assert.NoError(t, err)

	err = conn.Send("HMSET", redisKeyHeartbeat(ns, "1"),
		"heartbeat_at", time.Now().Add(-1*time.Hour).Unix(),
	)
	assert.NoError(t, err)

	err = conn.Send("HMSET", redisKeyHeartbeat(ns, "2"),
		"heartbeat_at", time.Now().Add(-1*time.Hour).Unix(),
		"job_names", "type1,type2",
	)
	assert.NoError(t, err)

	err = conn.Flush()
	assert.NoError(t, err)

	// Without this, the assertion for deadPools fails on GitHub CI. Could have
	// something to do with in-memory redis instance.
	time.Sleep(10 * time.Millisecond)

	// Test getting dead pool
	reaper := newDeadPoolReaper(ns, pool, []string{})
	deadPools, err := reaper.findDeadPools()
	assert.NoError(t, err)
	assert.Equal(t, map[string][]string{"2": {"type1", "type2"}}, deadPools)

	// Test requeueing jobs
	_, err = conn.Do("lpush", redisKeyJobsInProgress(ns, "1", "type1"), "foo")
	assert.NoError(t, err)
	_, err = conn.Do("lpush", redisKeyJobsInProgress(ns, "2", "type1"), "foo")
	assert.NoError(t, err)

	// Ensure 0 jobs in jobs queue
	jobsCount, err := redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)

	// Ensure 1 job in inprogress queue for each job
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "1", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Reap. Ensure job 2 is requeued but not job 1
	err = reaper.reap()
	assert.NoError(t, err)

	// Ensure 1 jobs in jobs queue
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobs(ns, "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Ensure 1 job in inprogress queue for 1
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "1", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 1, jobsCount)

	// Ensure 0 jobs in inprogress queue for 2
	jobsCount, err = redis.Int(conn.Do("llen", redisKeyJobsInProgress(ns, "2", "type1")))
	assert.NoError(t, err)
	assert.Equal(t, 0, jobsCount)
}

func TestDeadPoolReaperWithWorkerPools(t *testing.T) {
	pool := newTestPool(t)
	ns := "work"
	job1 := "job1"
	stalePoolID := "aaa"
	cleanKeyspace(ns, pool)
	// test vars
	expectedDeadTime := 5 * time.Millisecond

	// create a stale job with a heartbeat
	conn := pool.Get()
	defer conn.Close()
	_, err := conn.Do("SADD", redisKeyWorkerPools(ns), stalePoolID)
	assert.NoError(t, err)
	_, err = conn.Do("LPUSH", redisKeyJobsInProgress(ns, stalePoolID, job1), `{"sleep": 10}`)
	assert.NoError(t, err)
	jobTypes := map[string]*jobType{"job1": nil}
	staleHeart := newWorkerPoolHeartbeater(ns, pool, stalePoolID, jobTypes, 1, []string{"id1"})
	staleHeart.start()

	// should have 1 stale job and empty job queue
	assert.EqualValues(t, 1, listSize(pool, redisKeyJobsInProgress(ns, stalePoolID, job1)))
	assert.EqualValues(t, 0, listSize(pool, redisKeyJobs(ns, job1)))

	// setup a worker pool and start the reaper, which should restart the stale job above
	wp := setupTestWorkerPool(pool, ns, job1, 1, JobOptions{Priority: 1})
	wp.deadPoolReaper = newDeadPoolReaper(wp.namespace, wp.pool, []string{"job1"})
	wp.deadPoolReaper.deadTime = expectedDeadTime
	wp.deadPoolReaper.start()

	// sleep long enough for staleJob to be considered dead
	time.Sleep(expectedDeadTime * 2)

	// now we should have 1 job in queue and no more stale jobs
	assert.EqualValues(t, 1, listSize(pool, redisKeyJobs(ns, job1)))
	assert.EqualValues(t, 0, listSize(pool, redisKeyJobsInProgress(ns, wp.workerPoolID, job1)))
	staleHeart.stop()
	wp.deadPoolReaper.stop()
}

func TestDeadPoolReaperCleanStaleLocks(t *testing.T) {
	pool := newTestPool(t)
	ns := "work"
	cleanKeyspace(ns, pool)

	conn := pool.Get()
	defer conn.Close()
	job1, job2 := "type1", "type2"
	jobNames := []string{job1, job2}
	workerPoolID1, workerPoolID2 := "1", "2"
	lock1 := redisKeyJobsLock(ns, job1)
	lock2 := redisKeyJobsLock(ns, job2)
	lockInfo1 := redisKeyJobsLockInfo(ns, job1)
	lockInfo2 := redisKeyJobsLockInfo(ns, job2)

	// Create redis data
	var err error
	err = conn.Send("SET", lock1, 3)
	assert.NoError(t, err)
	err = conn.Send("SET", lock2, 1)
	assert.NoError(t, err)
	err = conn.Send("HSET", lockInfo1, workerPoolID1, 1) // workerPoolID1 holds 1 lock on job1
	assert.NoError(t, err)
	err = conn.Send("HSET", lockInfo1, workerPoolID2, 2) // workerPoolID2 holds 2 locks on job1
	assert.NoError(t, err)
	err = conn.Send("HSET", lockInfo2, workerPoolID2, 2) // test that we don't go below 0 on job2 lock
	assert.NoError(t, err)
	err = conn.Flush()
	assert.NoError(t, err)

	reaper := newDeadPoolReaper(ns, pool, jobNames)
	// clean lock info for workerPoolID1
	err = reaper.cleanStaleLockInfo(workerPoolID1, jobNames)
	assert.NoError(t, err)
	assert.EqualValues(t, 2, getInt64(pool, lock1))   // job1 lock should be decr by 1
	assert.EqualValues(t, 1, getInt64(pool, lock2))   // job2 lock is unchanged
	v, _ := conn.Do("HGET", lockInfo1, workerPoolID1) // workerPoolID1 removed from job1's lock info
	assert.Nil(t, v)

	// now clean lock info for workerPoolID2
	err = reaper.cleanStaleLockInfo(workerPoolID2, jobNames)
	assert.NoError(t, err)
	// both locks should be at 0
	assert.EqualValues(t, 0, getInt64(pool, lock1))
	assert.EqualValues(t, 0, getInt64(pool, lock2))
	// worker pool ID 2 removed from both lock info hashes
	v, err = conn.Do("HGET", lockInfo1, workerPoolID2)
	assert.Nil(t, v)
	v, err = conn.Do("HGET", lockInfo2, workerPoolID2)
	assert.Nil(t, v)
}