def get_nm_and_check(self,host,port):
url = "http://%s:%s/dfshealth.jsp" % (host,port)
html = util.get_http(url,5)
if html == None :
self.update_result("ERROR",u"不能连接到%s:%s。" % (host,port) )
return ("",0)
#get ha_state tid
ha_state = self.get_ha_state(html)
tid = self.get_tid(html)
#get storge
dirs = self.get_storge_dir(html)
for (dir,dir_state) in dirs.items():
if dir_state != "Active":
self.update_result("ERROR",u"目录 %s 状态为 %s。" % (dir,dir_state) )
#get live datanode status
# {"name","pcremaining","volfails"}
url = "http://%s:%s/dfsnodelist.jsp?whatNodes=LIVE" % (host,port)
html = util.get_http(url,5)
if html == None :
self.update_result("ERROR",u"不能连接到%s:%s。" % (host,port) )
return (ha_state,tid)
tables = self.get_live_dn_table(html)
for row in tables :
if row["volfails"] > 0 :
self.update_result("ERROR", u"机器 %s 检测到 %d 个损坏的卷。" % (row["name"],row["volfails"]) )
if row["last_contact"] > 30 :
self.update_result("ERROR", u"机器 %s 检测丢失心跳 %d秒。" % (row["name"],row["last_contact"]) )
return (ha_state,tid)
def check_nn_active(host, port):
url = "http://%s:%s/dfshealth.jsp" % (host,port)
html = util.get_http(url,5)
if not html: return False
ha_state = ''
pattern = re.compile("<h1>NameNode '.*?' \((.*?)\)</h1>",re.S)
m = pattern.search(html)
if m: ha_state = m.group(1)
return ha_state == 'active'
def get_nm_and_check(self, host, port):
url = "http://%s:%s/dfshealth.jsp" % (host, port)
html = util.get_http(url, 5)
if html == None:
self.update_result("%s(connect nmweb error)" % host,
u"不能连接到%s:%s。" % (host, port))
return ("", 0)
#get ha_state tid
ha_state = self.get_ha_state(html)
tid = self.get_tid(html)
#get storge
dirs = self.get_storge_dir(html)
for (dir, dir_state) in dirs.items():
if dir_state != "Active":
self.update_result("%s(storge error)" % (host, dir),
u"目录 %s 状态为 %s。" % (dir, dir_state))
#get live datanode status
# {"name","pcremaining","volfails"}
url = "http://%s:%s/dfsnodelist.jsp?whatNodes=LIVE" % (host, port)
html = util.get_http(url, 5)
if html == None:
self.update_result("%s(connect nmweb error)" % host,
u"不能连接到%s:%s。" % (host, port))
return (ha_state, tid)
tables = self.get_live_dn_table(html)
for row in tables:
if row["volfails"] > 0:
self.update_result(
"%s(datanode error Vol)" % row["name"],
u"机器 %s 检测到 %d 个损坏的卷。" % (row["name"], row["volfails"]))
if row["last_contact"] > 30:
self.update_result(
"%s(datanode lost heartbeat)" % row["name"],
u"机器 %s 检测丢失心跳 %d秒。" % (row["name"], row["last_contact"]))
return (ha_state, tid)
def getJobCounter(self,jobid):
"""
从appid的对应的counter的网页上截取信息.从restapi获取的不全,缺少data-local等的信息
"""
url = ("http://%s:%s/jobhistory/jobcounters/%s"
% (self.hshost,self.hsport,jobid))
html = util.get_http(url)
if not html:
return None
keys = ["DATA_LOCAL_MAPS","RACK_LOCAL_MAPS",
"FILE_BYTES_READ","FILE_BYTES_WRITTEN",
"HDFS_BYTES_READ","HDFS_BYTES_WRITTEN"]
counters = {}
for key in keys:
counters[key] = self.getCounterFromHtml(html,key)
return counters
def getJobCounter(self, jobid):
"""
从appid的对应的counter的网页上截取信息.从restapi获取的不全,缺少data-local等的信息
"""
url = ("http://%s:%s/jobhistory/jobcounters/%s" %
(self.hshost, self.hsport, jobid))
html = util.get_http(url)
if not html:
return None
keys = [
"DATA_LOCAL_MAPS", "RACK_LOCAL_MAPS", "FILE_BYTES_READ",
"FILE_BYTES_WRITTEN", "HDFS_BYTES_READ", "HDFS_BYTES_WRITTEN"
]
counters = {}
for key in keys:
counters[key] = self.getCounterFromHtml(html, key)
return counters