This blog is about
monitoring the Hadoop metrics such as DFS, MAPRED, JVM, RPC and
UGI using the Ganglia Monitoring Tool. I assume that the
readers of blog have prior knowledge of Ganglia and Hadoop
technology. To integrate the Ganglia with Hadoop you need to configure hadoop-metrics.properties file of
hadoop located inside the hadoop conf folder. In this configuration file you
need to configure the server address of ganglia gmetad, period for sending
metrics data and ganglia context class name. The format and name of hadoop
metrics properties file is different for various hadoop versions.
- For Hadoop 0.20.x, 0.21.0 and 0.22.0 versions, the file name is hadoop-metrics.properties.
- For Hadoop 1.x.x and 2.x.x versions, the file name is hadoop-metrics2.properties.
The ganglia context
class name also differs with version change of ganglia, for detailed
information about Ganglia Context class you can read from GangliaContext.
Procedure of
configuring the hadoop metrics properties file:
1. Configuration for
2.x.x versions: In
such hadoop versions the metrics properties file is located inside
the $HADOOP_HOME/etc/hadoop/
folder. Configure the hadoop-metrics2.properties file using the code as shown below:
namenode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
namenode.sink.ganglia.period=10
namenode.sink.ganglia.servers=gmetad_server_ip:8649
datanode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
datanode.sink.ganglia.period=10
datanode.sink.ganglia.servers=gmetad_server_ip:8649
resourcemanager.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
resourcemanager.sink.ganglia.period=10
resourcemanager.sink.ganglia.servers=gmetad_server_ip:8649
nodemanager.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
nodemanager.sink.ganglia.period=10
nodemanager.sink.ganglia.servers=gmetad_server_ip:8649
namenode.sink.ganglia.period=10
namenode.sink.ganglia.servers=gmetad_server_ip:8649
datanode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
datanode.sink.ganglia.period=10
datanode.sink.ganglia.servers=gmetad_server_ip:8649
resourcemanager.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
resourcemanager.sink.ganglia.period=10
resourcemanager.sink.ganglia.servers=gmetad_server_ip:8649
nodemanager.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
nodemanager.sink.ganglia.period=10
nodemanager.sink.ganglia.servers=gmetad_server_ip:8649
2.
Configuration for 1.x.x
versions: In
such hadoop versions the metrics properties file is located inside the $HADOOP_HOME/conf/ folder.
Configure the hadoop-metrics2.properties file using the code as shown below:
namenode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
namenode.sink.ganglia.period=10
namenode.sink.ganglia.servers=gmetad_server_ip:8649
datanode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
datanode.sink.ganglia.period=10
datanode.sink.ganglia.servers=gmetad_server_ip:8649
jobtracker.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
jobtracker.sink.ganglia.period=10
jobtracker.sink.ganglia.servers=gmetad_server_ip:8649
tasktracker.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
tasktracker.sink.ganglia.period=10
tasktracker.sink.ganglia.servers=gmetad_server_ip:8649
namenode.sink.ganglia.period=10
namenode.sink.ganglia.servers=gmetad_server_ip:8649
datanode.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
datanode.sink.ganglia.period=10
datanode.sink.ganglia.servers=gmetad_server_ip:8649
jobtracker.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
jobtracker.sink.ganglia.period=10
jobtracker.sink.ganglia.servers=gmetad_server_ip:8649
tasktracker.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
tasktracker.sink.ganglia.period=10
tasktracker.sink.ganglia.servers=gmetad_server_ip:8649
maptask.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
maptask.sink.ganglia.period=10
maptask.sink.ganglia.servers=gmetad_server_ip:8649
maptask.sink.ganglia.period=10
maptask.sink.ganglia.servers=gmetad_server_ip:8649
reducetask.sink.ganglia.class=org.apache.hadoop.metrics2.sink.ganglia.GangliaSink31
reducetask.sink.ganglia.period=10
reducetask.sink.ganglia.servers=gmetad_server_ip:8649
reducetask.sink.ganglia.period=10
reducetask.sink.ganglia.servers=gmetad_server_ip:8649
3.
Configuration for 0.20.x,
0.21.0 and 0.22.0 versions: In such hadoop versions the metrics
properties file is located inside the $HADOOP_HOME/conf/
folder. Configure the hadoop-metrics.properties file using the code as shown below:
dfs.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
dfs.period=10
dfs.servers=gmetad_server_ip:8649
mapred.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
mapred.period=10
mapred.servers=gmetad_server_ip:8649
jvm.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
jvm.period=10
jvm.servers=gmetad_server_ip:8649
rpc.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
rpc.period=10
rpc.servers=gmetad_server_ip:8649
dfs.period=10
dfs.servers=gmetad_server_ip:8649
mapred.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
mapred.period=10
mapred.servers=gmetad_server_ip:8649
jvm.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
jvm.period=10
jvm.servers=gmetad_server_ip:8649
rpc.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
rpc.period=10
rpc.servers=gmetad_server_ip:8649
ugi.class=org.apache.hadoop.metrics.ganglia.GangliaContext31
ugi.period=10
ugi.servers=gmetad_server_ip:8649
ugi.period=10
ugi.servers=gmetad_server_ip:8649
The above
configuration is for the unicast
mode of Ganglia. However, if you are running Ganglia in multicast mode then you need to use the multicast address in place of gmetad_server_ip
in the configuration file. Once you have applied the above changes, then
you need to restart the gmetad and gmond services of Ganglia on the nodes. You also
need to restart Hadoop services if they are running. Once you are done with restarting the services, the Ganglia UI
displays the Hadoop graphs. Initially Ganglia UI does
not show graphs for the jobs, they will appear on UI only after submitting a
job in Hadoop.
I'm new to hadoop and trying to monitor the working of a multi node cluster using ganglia, The setup of the daemons is done on all nodes however hadoop metrics graphs are available on the master node but not on the slaves. Do the graphs on the master include the slave metrics as well? Any help would be appreciated.
ReplyDeleteHi Anish,
DeleteYou need to create the hadoop metrics properties file on each node of cluster under hadoop configuration folder along with the configuration of file as shown above.
After configuration you need to start/restart the hadoop services based on the current service status.
And the graphs on master node does not include the graphs of slave nodes. The slave node itself shows the hadoop metric graphs on node level view of ganglia UI,
Please let me know if you still face the same issue.
I recently came across your blog on hadoop and have been reading along. I thought I would leave my first comment. I don’t know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.
ReplyDeleteHadoop Training in hyderabad
This information you provided in the blog that was really unique I love it!!, Thanks for sharing such a great blog..Keep posting..
ReplyDeleteBig Data Training in Chennai
This post is great. The configuration for Hadoop 2 works fine and your explanation is simple and helpful.
ReplyDeleteThanks.
I had got a clear idea of what the blog is all about. Kindly continue doing more.
ReplyDeleteTOEFL Coaching in Chennai
Best TOEFL Coaching Institute in Chennai
TOEFL Classes in Chennai
TOEFL Training in Chennai
IELTS Classes in Mumbai
Spoken English Classes in JP Nagar
Spoken English Class in Thiruvanmiyur
Spoken English Classes in Chennai
English Speaking Classes in Mumbai
I am really enjoying reading your well written articles.
ReplyDeleteIt looks like you spend a lot of effort and time on your blog.
I have bookmarked it and I am looking forward to reading new articles.Keep up the good work..
RPA Training in Chennai
RPA Classes in Chennai
Blue Prism Training in Chennai
Ethical Hacking Training in Chennai
Cloud Computing Training in Chennai
RPA Training in T Nagar
RPA Training in Porur
Mmorpg Oyunlar
ReplyDeleteınstagram takipci satin al
tiktok jeton hilesi
tiktok jeton hilesi
antalya saç ekimi
referans kimliği nedir
İnstagram Takipçi Satın Al
Mt2 Pvp Serverler
takipci
perde modelleri
ReplyDeletemobil onay
MOBİL ODEME BOZDURMA
Nftnasilalinir
ANKARA EVDEN EVE NAKLİYAT
trafik sigortasi
dedektor
Web site kurma
aşk kitapları
SMM PANEL
ReplyDeletesmm panel
İs İlanlari
İNSTAGRAM TAKİPÇİ SATIN AL
HİRDAVATCİBURADA.COM
beyazesyateknikservisi.com.tr
SERVİS
tiktok jeton hilesi
pendik bosch klima servisi
ReplyDeletekartal beko klima servisi
beykoz lg klima servisi
beykoz arçelik klima servisi
çekmeköy mitsubishi klima servisi
ataşehir mitsubishi klima servisi
maltepe bosch klima servisi
ümraniye mitsubishi klima servisi
beykoz bosch klima servisi