Fri Aug 17, 2018 11:04 pm
Login Register Lost Password? Contact Us

Mythor not starting after 6.0.0-x Install

Post questions specific to installation or configuration for the HPCC Systems platform

Thu Jun 23, 2016 8:36 pm Change Time Zone


On recent installs of both 6.0.0-1 and 6.0.0-2, I've encountered identical problems preventing the startup of mythor, and affecting it's connections to the mythor slaves.
After a "sudo service hpcc-init restart" call, all components start successfully, except mythor. I've attached the relevant log files, but I believe the originating issues are in /var/log/HPCCSystems/frunssh/*.log, reading:
Code: Select all
ERROR: /var/lib/jenkins2/workspace/CE-Candidate-6.0.0-1/CE/ubuntu-14.04amd64/HPCC-Platform/services/runagent/frunssh.cpp(73) : frunssh : Failed to open slaves file /var/lib/HPCCSystems/mythor/slaves

and in /var/log/HPCCSystems/mythor/init_mythor_<timestamp>.log:
Code: Select all
2016-06-23T20:12:08: Killing slaves
2016-06-23T20:12:08: Error 255 in frunssh
2016-06-23T20:12:08: Please check /var/log/frunssh for more details
2016-06-23T20:12:08: Stopping mythor

I've also noticed that the file /var/lib/HPCCSystems/mythor/slaves does not exist, but a file /var/lib/HPCCSystems does. On a whim, changing the name of this "uslaves" file to "slaves" changes the output of the "init_mythor..." log to:
Code: Select all
1: starting (0 of 2 finished)
0: starting (0 of 2 finished)
Results: (2 of 2 finished)
1: starting (0 of 2 finished)
0: starting (0 of 2 finished)
Results: (2 of 2 finished)

Any ideas?

(38.17 KiB) Downloaded 120 times
Posts: 3
Joined: Mon Feb 01, 2016 2:03 am

Mon Jul 04, 2016 12:07 pm Change Time Zone


the slaves file will not normally exist after Thor has stopped or failed to start, since it is a temporary file that's cleared up on exit.

In start_slaves_01_22_2015_13_26_44.log, I notice:
/opt/HPCCSystems/bin/start_slaves: line 77: rsync: command not found
cat: /var/lib/HPCCSystems/mythor/thorgroup.slave: No such file or directory

I suspect that's the problem. It's expecting rsync to be installed, but it appears not to be. The use of rsync here isn't new to 6.0.0-x though as far as I know.
Community Advisory Board Member
Community Advisory Board Member
Posts: 70
Joined: Tue Jul 19, 2011 12:58 pm

Mon Jul 04, 2016 12:16 pm Change Time Zone

I think rsync is installed on most distros by default, but to fix this you'll need to install it.

On CentOS you can install with:
sudo yum -y install rsync

On Ubuntu you can install with:
sudo apt-get -y install rsync
Community Advisory Board Member
Community Advisory Board Member
Posts: 70
Joined: Tue Jul 19, 2011 12:58 pm

Wed Jul 06, 2016 4:11 pm Change Time Zone

Hi demills,

Any chance you can check on the status of rsync on your system? If you can't find the rsync package in your $PATH, but the package manager says it's installed, can you please print out your $PATH for us? There could possibly be an issue with your environment variables.

Rsync should be installed as a dependency when you install the platform package. Also, did you install the package (on the slave) the standard way, with 'dpkg -i hpccsystems-platform..., apt-get install -f'?

Best regards,

Michael Gardner
Posts: 13
Joined: Tue Jan 20, 2015 9:30 pm

Mon Oct 24, 2016 9:23 pm Change Time Zone

Dear all,

I have encounted the same issue. After "sudo /opt/HPCCSystems/sbin/ -a hpcc-init start" called, all components can be started successfully, except mythor. Then I called "sudo service hpcc-init restart" on the master node and I got:
Code: Select all
ubuntu@ubuntu01:/etc/HPCCSystems$ sudo service hpcc-init restart
Stopping entire system for a full restart
Stopping mythor...             [   OK    ]
Stopping mysasha...            [   OK    ]
Stopping myesp...              [   OK    ]
Stopping myeclscheduler...     [   OK    ]
Stopping myeclccserver...      [   OK    ]
Stopping myeclagent...         [   OK    ]
Stopping mydfuserver...        [   OK    ]
Stopping mydali...             [   OK    ]
Starting the entire System
Starting mydali ...            [   OK    ]
Starting mydfuserver ...       [   OK    ]
Starting myeclagent ...        [   OK    ]
Starting myeclccserver ...     [   OK    ]
Starting myeclscheduler ...    [   OK    ]
Starting myesp ...             [   OK    ]
Starting mysasha ...           [   OK    ]
Starting mythor ...            [ FAILED  ]

I checked "/var/log/HPCCSystems/mythor/init_mythor_<timestamp>.log" and it shows:
Code: Select all
2016-10-24T21:09:33: Starting mythor
2016-10-24T21:09:33: removing any previous sentinel file
2016-10-24T21:09:33: Ensuring a clean working environment ...
2016-10-24T21:09:33: Killing slaves
2016-10-24T21:09:35: Error 255 in frunssh
2016-10-24T21:09:35: Please check /var/log/HPCCSystems/frunssh for more details
2016-10-24T21:09:35: Stopping mythor

and in "/var/log/HPCCSystems/frunssh/*.log" it shows:
Code: Select all
1: ssh(0):
2: ssh(0):
3: ssh(0):
ERROR: /mnt/disk1/jenkins/workspace/CE-Candidate-6.0.6-1/CE/ubuntu-14.04-amd64/HPCC-Platform/services/runagent/frunssh.cpp(84) : frunssh : [255: ]
ERROR: /mnt/disk1/jenkins/workspace/CE-Candidate-6.0.6-1/CE/ubuntu-14.04-amd64/HPCC-Platform/services/runagent/frunssh.cpp(84) : frunssh : [255: ]
ERROR: /mnt/disk1/jenkins/workspace/CE-Candidate-6.0.6-1/CE/ubuntu-14.04-amd64/HPCC-Platform/services/runagent/frunssh.cpp(84) : frunssh : [255: ]

The complete log files can be found in the attachment.

Please advise. Thank you.

(67.04 KiB) Downloaded 104 times
Posts: 27
Joined: Fri Jul 25, 2014 6:00 pm

Return to Installation

Who is online

Users browsing this forum: No registered users and 1 guest