Request to the Host Monitor failed. This may cause slow page responses.

[#1]

Hi All,

I my cluster was brought down to increase ram to all 8 node. Later when i started hadoop services i had issues with mysqld start which i fixed. now I am able to access cloudera manager webpage from namenode Ip. Its blank and i have this error

" Request to the Host Monitor failed. This may cause slow page responses. "  and

"Internal error while querying the Host Monitor "

I tried to restart and check log there is notihing on it. any suggestion is appreciated.

Regards,

Dharani

[#2]

Check the cloudera management services, is the Host Monitor running?
If not, check why in the logs under /var/log/cloudera-scm-firehose.

[#3]

Thanks Gautam for replying,

Host monitor is not running, and I am not able to locate the cloudera-scm-firehose.log file in any directory. I just have one for scm-server and scm-agent. 

[#4]

Please look up the directory /var/log/cloudera-scm-firehose (not
cloudera-scm-server) on the host where the Host Monitor is running. It
does not have to run on the same host as Cloudera Manager. If it was
running earlier, the log would have to be present somewhere.

[#5]

Thanks gautam, I had a look at firehose file and found below error in log

java.net.SocketException: Network is unreachable

ERROR com.cloudera.cmon.firehose.Main: Could not fetch descriptor after 5 tries, exiting.

WARN com.cloudera.cmon.firehose.Main: No descriptor fetched from http://dhcp-xxxxxxxxxxxxx on after 5 tries, sleeping.

Any suggestion on how to fix this ?

Regards

Dharani

[#6]

Hi, I am having the same issue.

---

@Dharani wrote:

Thanks gautam, I had a look at firehose file and found below error in log

java.net.SocketException: Network is unreachable

ERROR com.cloudera.cmon.firehose.Main: Could not fetch descriptor after 5 tries, exiting.

WARN com.cloudera.cmon.firehose.Main: No descriptor fetched from http://dhcp-xxxxxxxxxxxxx on after 5 tries, sleeping.

Any suggestion on how to fix this ?

Regards

Dharani

---

[#7]

Hi 

Check for the java -version, and if the jdk is installed, check the path setup in ~/.bash_profile. 

Once all are set. restart and check the cloudera-scm-agent. 

and go and refresh the weburl. 

Check it might work.

Thanks

[#8]

Hi @srigowri ,

Thank you for helping out in the community.  This thread is over 4 years old, so it is unlikely that the issue is still occurring.

Also note that the error message said that it could not fetch the descriptor.  This means that the Host Monitor had trouble retrieving cluster information from Cloudera Manager.  When this happens, Host Monitor will not be functional since it is not aware of the cluster layout.

The best course of action in this case is to look at what URL the Host Monitor is using to connect to Cloudera Manager and test with curl or a browser to see what happens when attempting to connect to the same URL.

This trouble can often occur when enabling TLS for Cloudera Manager.  If you enable TLS and restart CM, the Management Service (which includes Host Monitor) would need to be restarted via Cloudera Manager in order for it to obtain the new URL that includes "https".

Without knowing more about what happened 4 years ago, though, we can't know for sure.

Cheers,

Ben