Improvements

[jm66]

• Logging improvements
• _list_vsphere_volumes will go through VMs sitting in configured datastore (agent.yml) (speed things up)
• _get_all_vms uses PropertyCollector to retrieve VMs with name and config.hardware attributes only (speed things up)
• Support for pyvmomi 6.0.0 and python 2.7.10 introduced by flocker in version 1.10.1 on ubuntu 15.10
• Updating Documentation
• Fixes #7

[pratikvgupta]

Hi Jose

Your change looks really great. Thanks for fixing this. I was planning to use the same approach to fix this. Does this also fix #7 now? Were you able to deploy container now on your inventory?

[jm66]

Unfortunately no. Now that I've deployed this new code I can see clearly what's going on but for some reason I can't get the driver to attach the disk to the vm and then present it to the OS. I'm planning later on this week to re-install flocker and start from scratch. Maybe something odd is going on in between flocker-* services and the driver. I think at this point I'd need someone from flocker to help me out troubleshooting.

Anyway, I thought these changes would improve in some way the driver. Do you have access to a lab where the code can be tested?

Thanks, JM

[jm66]

Hi @pratikvgupta working now and times look not that bad:

2016-02-04 11:43:31,181 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 84.6229529381 seconds
2016-02-04 11:43:31,603 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 85.0442380905 seconds
2016-02-04 11:43:31,866 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 85.3068330288 seconds
2016-02-04 11:43:31,859 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 85.2997620106 seconds
2016-02-04 11:43:34,323 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 0.978564977646 seconds
2016-02-04 11:44:56,902 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 82.5556080341 seconds
2016-02-04 11:44:58,017 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 83.669426918 seconds
2016-02-04 11:44:58,325 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 83.9783930779 seconds
2016-02-04 11:44:58,591 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 84.2455000877 seconds
2016-02-04 11:45:06,620 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.45467805862 seconds
2016-02-04 11:46:20,731 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 70.2822551727 seconds
2016-02-04 11:46:21,006 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 70.5582768917 seconds
2016-02-04 11:46:21,134 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 70.6859121323 seconds
2016-02-04 11:46:28,802 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.00356698036 seconds
2016-02-04 11:47:12,947 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 41.3331410885 seconds
2016-02-04 11:47:13,816 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 42.2221980095 seconds
2016-02-04 11:47:21,256 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 0.951939105988 seconds
2016-02-04 11:47:37,789 - vsphere_blockdevice - DEBUG - _get_all_vms: Took 13.8123800755 seconds
2016-02-04 11:47:43,630 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.04235696793 seconds
2016-02-04 11:47:48,131 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 0.993250131607 seconds
2016-02-04 11:47:49,769 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.02728700638 seconds
2016-02-04 11:48:52,205 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.09965801239 seconds
2016-02-04 11:49:51,730 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.30625414848 seconds
2016-02-04 11:51:03,894 - vsphere_blockdevice - DEBUG - list_volumes: List volumes finished. Took 1.09758281708 seconds

Everything works pretty good so far. I've tried:

• docker container deployment with volume provided by flocker
• docker container deletion
• dataset deletion via flocker api