Yeah, those version dependencies can be a bit of a pain. I ended up hacking together a script <https://github.com/JonZeolla/Development/blob/master/Bash/setupMetron.sh> to make sure I pull down exactly the right versions, but it's only really helpful for CentOS 6.8 right now (because that's what I use). In theory it works, but I didn't finish my testing, so YMMV. Feel free to rip some of the guts out if it's helpful.
Jon On Mon, Nov 21, 2016 at 1:58 PM Otto Fowler <[email protected]> wrote: > Matt, > They have to be EXACTLY the versions. > > > > On November 21, 2016 at 13:27:49, Matt Foley ([email protected]) wrote: > > I had checked them, and they are all >= the specified requirement in > > https://github.com/apache/incubator-metron/tree/master/metron-deployment/vagrant/quick-dev-platform#user-content-prerequisites > : > Ansible 2.1.2.0 > Vagrant 1.8.6 > Virtualbox 5.0.24 > Python 2.7.12 > Maven 3.3.9 > > The only one significantly different is Ansible; the requirement is > 2.0.0.2. > Is Ansible 2.1 not compatible with Ansible 2.0? I let homebrew install the > latest version. > Thanks, > --Matt > > From: Otto Fowler <[email protected]> > Date: Saturday, November 19, 2016 at 5:18 AM > To: "[email protected]" <[email protected]>, > Matt Foley <[email protected]> > Subject: Re: problem with vagrant launch > > What versions of ansible, vagrant, virtualbox and python do you have? > > > > On November 18, 2016 at 16:34:08, Matt Foley ([email protected]) wrote: > Hi, > I'm trying to launch vanilla vagrant single-node test env for Metron on a > Mac, via > “vagrant up” in metron-deployment/vagrant/full-dev-platform > > The Ambari install all goes fine, then the Elastic Search install > but when it comes to > TASK [metron_elasticsearch_templates : Add Elasticsearch templates for > topologies] > it fails on all three (bro, snort, yaf indexes) with > "Status code was not [200]: An unknown error occurred: sendall() argument 1 > must be string or buffer, not dict" > > Launching from metron-deployment/vagrant/quick-dev-platform had the same > failure point. > > All three *_index.template files are present in the specified directory. > > Does anyone recognize this problem? Have people been running the vagrant > test env lately? > Any suggestions for debugging? > > Thanks, > --Matt > > Cmd line output: > … > TASK [metron_elasticsearch_templates : Wait for Index to Become Available] > ***** > ok: [node1] > > TASK [metron_elasticsearch_templates : Add Elasticsearch templates for > topologies] *** > failed: [node1] > > (item=/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/bro_index.template) > => {"content": "", "failed": true, "item": > > "/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/bro_index.template", > "msg": "Status code was not [200]: An unknown error occurred: sendall() > argument 1 must be string or buffer, not dict", "redirected": false, > "status": -1, "url": "http://node1:9200/_template/bro_index"} > failed: [node1] > > (item=/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/snort_index.template) > => {"content": "", "failed": true, "item": > > "/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/snort_index.template", > "msg": "Status code was not [200]: An unknown error occurred: sendall() > argument 1 must be string or buffer, not dict", "redirected": false, > "status": -1, "url": "http://node1:9200/_template/snort_index"} > failed: [node1] > > (item=/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/yaf_index.template) > => {"content": "", "failed": true, "item": > > "/Users/mfoley/projects/Metron/metron532/metron-deployment/roles/metron_elasticsearch_templates/files/es_templates/yaf_index.template", > "msg": "Status code was not [200]: An unknown error occurred: sendall() > argument 1 must be string or buffer, not dict", "redirected": false, > "status": -1, "url": "http://node1:9200/_template/yaf_index"} > > NO MORE HOSTS LEFT > ************************************************************* > to retry, use: --limit > > @/Users/mfoley/projects/Metron/metron532/metron-deployment/playbooks/metron_full_install.retry > > > PLAY RECAP > ********************************************************************* > node1 : ok=44 changed=12 unreachable=0 > failed=1 > > Ansible failed to complete successfully. Any error output should be > visible above. Please fix these errors and try again. > -- Jon Sent from my mobile device
