Add a src-side script to display a summary of host info.

Add a src-side script to display a summary of host info.

The idea being that this make it easier to diagnose and fix issues like http://crbug.com/439598.

Sample run: http://simonhatch0.mon.corp.google.com:8013/builders/Android%20Nexus7v2%20Perf/builds/337/steps/host_info/logs/json.output

The output from this run on the Nexus 7v2 bots looks like:
{
  "failures": [
    "build_versions"
  ],
  "host_info": {
    "devices": {
      "build_types": "userdebug/dev-keys",
      "build_versions": "MISMATCH",
      "build_versions_list": [
        "4.4.2/KOT49H/937116",
        "5.0/LRX21P/1570855"
      ],
      "device_names": "google/razor/flo",
      "devices": [
        "09275b25",
        "09e0cda5",
        "09e19954",
        "09e19f78",
        "09e28a6c",
        "0a38ca3e",
        "0a9f170a",
        "0a9f1772"
      ]
    },
    "free_disk_space": 543.9997482299805,
    "num_cpus": 40,
    "os_release": "3.13.0-44-generic",
    "os_system": "Linux",
    "processor": "x86_64",
    "python_path": "/usr/bin/python",
    "python_version": "2.7.6"
  },
  "valid": false
}

This can be enabled by adding this to testing/buildbot/chromium.perf.json:
{
  "Android Nexus7v2 Perf": {
    "scripts": [
      {
        "name": "host_info",
        "script": "host_info.py"
      }
    ]
  }
}

Needs to land before enabling (that CL enables src-side tests for perf bots): https://codereview.chromium.org/873403002/

BUG=443708, 454934
Committed: https://crrev.com/1730cecbc05928e93e2dec826a67139bf6765cc6
Cr-Commit-Position: refs/heads/master@{#316651}

[shatch, 2015-02-03 20:44:03]

simonhatch@chromium.org changed reviewers:
+ sullivan@chromium.org

[shatch, 2015-02-03 20:44:03]

Hey Annie, want to take a look at the output before I send this out?

Interesting to note this script found that the 7v2 bots are mostly running K,
with one stray one running L??

[sullivan, 2015-02-04 04:26:15]

On 2015/02/03 20:44:03, shatch wrote:
> Hey Annie, want to take a look at the output before I send this out?
> 
> Interesting to note this script found that the 7v2 bots are mostly running K,
> with one stray one running L??

Output looks great, has everything I would expect. And already found a bug to
boot :)

[shatch, 2015-02-04 15:43:56]

Patchset #2 (id:20001) has been deleted

[shatch, 2015-02-04 15:46:17]

simonhatch@chromium.org changed reviewers:
+ iannucci@chromium.org
- sullivan@chromium.org

[shatch, 2015-02-04 15:46:18]

phajdan is OOO this week, ptal

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
File testing/scripts/host_info.py (right):

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:122: 'host_info': host_info,
I'm not really clear what's allowed in the json output (if anything beyond
valid/failures?), so I tucked the host info in here for now so that it appears
on the waterfall.

[shatch, 2015-02-09 21:26:40]

simonhatch@chromium.org changed reviewers:
+ phajdan.jr@chromium.org

[shatch, 2015-02-09 21:26:41]

Adding phajdan since he's back this week.

[Paweł Hajdan Jr., 2015-02-10 08:34:12]

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
File testing/scripts/host_info.py (right):

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:34: assert False, 'OS %s not supported.' % os.name
I'd rather throw an exception than assert.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:83: 'device_names':
sorted(list(set([v.split(':')[0] for v in details]))),
Consider extracting the common logic to a helper (sorted, set, ...); looks like
the only thing that changes is the index.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:107: if host_info['num_cpus'] == 'unknown':
Why not put this into get_num_cpus?

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:111: if host_info['free_disk_space'] == 0:
Why not put this into get_free_disk_space?

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:122: 'host_info': host_info,
On 2015/02/04 15:46:18, shatch wrote:
> I'm not really clear what's allowed in the json output (if anything beyond
> valid/failures?), so I tucked the host info in here for now so that it appears
> on the waterfall.

Maybe prefix it with '_' to indicate it's a custom field.

[shatch, 2015-02-10 16:42:16]

Patchset #3 (id:60001) has been deleted

[shatch, 2015-02-10 16:42:58]

ptal

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
File testing/scripts/host_info.py (right):

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:34: assert False, 'OS %s not supported.' % os.name
On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> I'd rather throw an exception than assert.

Removed assert and just added to the failures list.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:83: 'device_names':
sorted(list(set([v.split(':')[0] for v in details]))),
On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> Consider extracting the common logic to a helper (sorted, set, ...); looks
like
> the only thing that changes is the index.

Done.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:107: if host_info['num_cpus'] == 'unknown':
On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> Why not put this into get_num_cpus?

Done.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:111: if host_info['free_disk_space'] == 0:
On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> Why not put this into get_free_disk_space?

Done.

https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
testing/scripts/host_info.py:122: 'host_info': host_info,
On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> On 2015/02/04 15:46:18, shatch wrote:
> > I'm not really clear what's allowed in the json output (if anything beyond
> > valid/failures?), so I tucked the host info in here for now so that it
appears
> > on the waterfall.
> 
> Maybe prefix it with '_' to indicate it's a custom field.

Done.

[shatch, 2015-02-17 15:28:50]

On 2015/02/10 16:42:58, shatch wrote:
> ptal
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> File testing/scripts/host_info.py (right):
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> testing/scripts/host_info.py:34: assert False, 'OS %s not supported.' %
os.name
> On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> > I'd rather throw an exception than assert.
> 
> Removed assert and just added to the failures list.
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> testing/scripts/host_info.py:83: 'device_names':
> sorted(list(set([v.split(':')[0] for v in details]))),
> On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> > Consider extracting the common logic to a helper (sorted, set, ...); looks
> like
> > the only thing that changes is the index.
> 
> Done.
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> testing/scripts/host_info.py:107: if host_info['num_cpus'] == 'unknown':
> On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> > Why not put this into get_num_cpus?
> 
> Done.
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> testing/scripts/host_info.py:111: if host_info['free_disk_space'] == 0:
> On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> > Why not put this into get_free_disk_space?
> 
> Done.
> 
>
https://codereview.chromium.org/899693002/diff/40001/testing/scripts/host_inf...
> testing/scripts/host_info.py:122: 'host_info': host_info,
> On 2015/02/10 08:34:12, Paweł Hajdan Jr. (OOO) wrote:
> > On 2015/02/04 15:46:18, shatch wrote:
> > > I'm not really clear what's allowed in the json output (if anything beyond
> > > valid/failures?), so I tucked the host info in here for now so that it
> appears
> > > on the waterfall.
> > 
> > Maybe prefix it with '_' to indicate it's a custom field.
> 
> Done.

ping

[Paweł Hajdan Jr., 2015-02-17 16:07:01]

https://codereview.chromium.org/899693002/diff/80001/testing/scripts/host_inf...
File testing/scripts/host_info.py (right):

https://codereview.chromium.org/899693002/diff/80001/testing/scripts/host_inf...
testing/scripts/host_info.py:119: 'valid': bool(not failures),
This is _valid_ , not e.g. "successful". Looks like it should always be True.

[shatch, 2015-02-17 16:49:22]

https://codereview.chromium.org/899693002/diff/80001/testing/scripts/host_inf...
File testing/scripts/host_info.py (right):

https://codereview.chromium.org/899693002/diff/80001/testing/scripts/host_inf...
testing/scripts/host_info.py:119: 'valid': bool(not failures),
On 2015/02/17 16:07:01, Paweł Hajdan Jr. wrote:
> This is _valid_ , not e.g. "successful". Looks like it should always be True.

Done.

[Paweł Hajdan Jr., 2015-02-17 17:29:08]

LGTM

[shatch, 2015-02-17 20:39:53]

The CQ bit was checked by simonhatch@chromium.org

[commit-bot: I haz the power, 2015-02-17 20:40:24]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/899693002/100001

[commit-bot: I haz the power, 2015-02-17 21:05:49]

Committed patchset #4 (id:100001)

[commit-bot: I haz the power, 2015-02-17 21:06:32]

Patchset 4 (id:??) landed as
https://crrev.com/1730cecbc05928e93e2dec826a67139bf6765cc6
Cr-Commit-Position: refs/heads/master@{#316651}