def resourcemanager_web(host,port):
'''
检查以下事项:
1. 不健康的node 及其对应的report
'''
state = "OK"
msg = u""
#state list:DECOMMISSIONED,LOST,NEW,REBOOTED,RUNNING,UNHEALTHY,
url = "http://%s:%s/ws/v1/cluster/nodes?state=unhealthy" % (host,port)
nodes = util.get_http_json(url,5)
if nodes != None :
if nodes["nodes"] != None :
for node in nodes["nodes"]["node"] :
state = "ERROR"
msg += u"检测到不健康机器 %s 健康报告 %s." % (node["nodeHostName"],node["healthReport"])
else:
return ("ERROR",u"不能连接到RM %s:%s" % (host,port) )
url = "http://%s:%s/ws/v1/cluster/nodes?state=lost" % (host,port)
nodes = util.get_http_json(url,5)
if nodes != None :
if nodes["nodes"] != None :
for node in nodes["nodes"]["node"] :
state = "ERROR"
ts = time.localtime(node['lastHealthUpdate']/1000)
date_time = time.strftime("%Y-%m-%d %H:%M:%S",ts)
msg += u"检测到丢失的机器 %s 丢失时间 %s." % (node["nodeHostName"],date_time)
else:
return ("ERROR",u"不能连接到RM %s:%s" % (host,port) )
return (state,msg)
def getJobAllTask(self, jobid):
url = "http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks" % (self.hshost, self.hsport, jobid)
tasks = util.get_http_json(url)
if not tasks or not tasks.has_key("tasks") or not tasks["tasks"].has_key("task"):
return
for task in tasks["tasks"]["task"]:
taskId = task["id"]
url = "http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts" % (
self.hshost,
self.hsport,
jobid,
taskId,
)
attempts = util.get_http_json(url)
if (
not attempts
or not attempts.has_key("taskAttempts")
or not attempts["taskAttempts"].has_key("taskAttempt")
):
return
for attempt in attempts["taskAttempts"]["taskAttempt"]:
attemptId = attempt["id"]
url = "http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts/%s/counters" % (
self.hshost,
self.hsport,
jobid,
taskId,
attemptId,
)
attemptCounter = util.get_http_json(url)
self.updateWithAttempt(attempt, attemptCounter)
def getAppList(self):
url = "http://%s:%s/ws/v1/cluster/apps?finishedTimeBegin=%d&finishedTimeEnd=%d" % (
self.rmhost,
self.rmport,
(self.recordTime * 1000),
(self.recordTime + self.interval) * 1000,
)
return util.get_http_json(url)
def getJobAllTask(self,jobid):
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks"
% (self.hshost,self.hsport,jobid))
tasks = util.get_http_json(url)
if not tasks or not tasks.has_key('tasks') or not tasks['tasks'].has_key('task') :
return
for task in tasks['tasks']['task']:
taskId = task['id']
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts"
% (self.hshost,self.hsport,jobid,taskId))
attempts = util.get_http_json(url)
if not attempts or not attempts.has_key('taskAttempts') \
or not attempts['taskAttempts'].has_key('taskAttempt') :
return
for attempt in attempts['taskAttempts']['taskAttempt']:
attemptId = attempt['id']
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts/%s/counters"
% (self.hshost,self.hsport,jobid,taskId,attemptId))
attemptCounter = util.get_http_json(url)
self.updateWithAttempt(attempt,attemptCounter)
def resourcemanager_web(host,port):
'''
检查以下事项:
1. 不健康的node 及其对应的report
'''
msg = u""
alarm_list = []
#state list:DECOMMISSIONED,LOST,NEW,REBOOTED,RUNNING,UNHEALTHY,
url = "http://%s:%s/ws/v1/cluster/nodes?state=unhealthy" % (host,port)
nodes = util.get_http_json(url,5)
if nodes != None :
if nodes["nodes"] != None :
for node in nodes["nodes"]["node"] :
key_word = "%s(rm unhealty node)" % node["nodeHostName"]
msg = u"检测到不健康机器 %s 健康报告 %s." % (node["nodeHostName"],node["healthReport"])
alarm_list.append({"key_word":key_word,"msg":msg})
else:
key_word = "%s(connect rm error)" % host
msg = u"不能连接到RM %s:%s" % (host,port)
alarm_list.append({"key_word":key_word,"msg":msg})
return alarm_list
url = "http://%s:%s/ws/v1/cluster/nodes?state=lost" % (host,port)
nodes = util.get_http_json(url,5)
if nodes != None :
if nodes["nodes"] != None :
for node in nodes["nodes"]["node"] :
ts = time.localtime(node['lastHealthUpdate']/1000)
date_time = time.strftime("%Y-%m-%d %H:%M:%S",ts)
msg = u"检测到丢失的机器 %s 丢失时间 %s." % (node["nodeHostName"],date_time)
key_word = "%s(rm lost node)" % node["nodeHostName"]
alarm_list.append({"key_word":key_word,"msg":msg})
else:
key_word = "%s(connect rm error)" % host
msg = u"不能连接到RM %s:%s" % (host,port)
alarm_list.append({"key_word":key_word,"msg":msg})
return alarm_list
return alarm_list
def getJobAllTask(self, jobid):
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks" %
(self.hshost, self.hsport, jobid))
tasks = util.get_http_json(url)
if not tasks or not tasks.has_key(
'tasks') or not tasks['tasks'].has_key('task'):
return
for task in tasks['tasks']['task']:
taskId = task['id']
url = (
"http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts"
% (self.hshost, self.hsport, jobid, taskId))
attempts = util.get_http_json(url)
if not attempts or not attempts.has_key('taskAttempts') \
or not attempts['taskAttempts'].has_key('taskAttempt') :
return
for attempt in attempts['taskAttempts']['taskAttempt']:
attemptId = attempt['id']
url = (
"http://%s:%s/ws/v1/history/mapreduce/jobs/%s/tasks/%s/attempts/%s/counters"
% (self.hshost, self.hsport, jobid, taskId, attemptId))
attemptCounter = util.get_http_json(url)
self.updateWithAttempt(attempt, attemptCounter)
def getMetrics(self):
url = ("http://%s:%s/ws/v1/cluster/metrics" % (self.rmhost,self.rmport))
return util.get_http_json(url)
def getJobHistory(self,jobid):
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s"
% (self.hshost,self.hsport,jobid))
return util.get_http_json(url)
def getAppList(self):
url = (
"http://%s:%s/ws/v1/cluster/apps?finishedTimeBegin=%d&finishedTimeEnd=%d"
% (self.rmhost, self.rmport, (self.recordTime * 1000),
(self.recordTime + self.interval) * 1000))
return util.get_http_json(url)
def getMetrics(self):
url = ("http://%s:%s/ws/v1/cluster/metrics" %
(self.rmhost, self.rmport))
return util.get_http_json(url)
def getJobHistory(self, jobid):
url = ("http://%s:%s/ws/v1/history/mapreduce/jobs/%s" %
(self.hshost, self.hsport, jobid))
return util.get_http_json(url)
