paasta metastatus should operate on pools

[EvanKrall]

Maybe we should actually make pool an argument to paasta metastatus so that you can look at a specific pool (does it ever make sense to consider the utilization of all pools added together? Seems like you want to know each pool individually).

Alternatively, paasta metastatus -vvv could print out the pool for each slave, letting you easily grep.

[Rob-Johnson]

you can do this, though metastatus is a bit more generic. The -g option to metastatus accepts an arbitrary attribute, and shows you usage per attribute value in the cluster. You can use this to group by pools.

robj@dev4-uswest1cdevc:~/yelpsoa-configs master % paasta metastatus -c norcal-prod -g pool -vv
Cluster: norcal-prod
Dashboards:
  Chronos RO:        http://chronos-norcal-prod.yelpcorp.com (LDAP Login)
  Mesos:             http://mesos.paasta-norcal-prod.yelp (Requires Prod Proxy, see http://y/paasta-prod-dashboards)
  To run metastatus: paasta metastatus -c norcal-prod (requires prod access)
  Smartstack:        https://systems-sfo2.yelpcorp.com/smartstack (Requires Prod Proxy, see http://y/paasta-prod-dashboards)
  Signalfx:          https://app.signalfx.com/#/dashboard/CYJklh9AYAo?sources%5B%5D=superregion:norcal-prod
  Marathon RO:       http://marathon-norcal-prod.yelpcorp.com (LDAP Login)
Master paasta_tools version: 0.60.2
Mesos Status: OK
  Quorum: masters: 3 configured quorum: 2 
  Frameworks:
    Framework: chronos count: 1
    Framework: marathon count: 1
  CPUs: 6801.19 / 8705 in use (78.13%)
  Memory: 16435.79 / 37706.42GB in use (43.59%)
  Disk: 6456.47 / 95742.16GB in use (6.74%)
  Tasks: running: 4155 staging: 2 starting: 0
  Slaves: active: 369 inactive: 0
  Connected Frameworks: expected: 2 actual: 2
  Disconnected Frameworks: expected: 0 actual: 0
  Inactive Frameworks: expected: 0 actual: 0
  Resources Grouped by pool
    Pool       CPU (used/total)               RAM (used/total)      Disk (used/total)
    cep159     7.0/210.0 (3.33%)              13.7G/342.1G (4.00%)  7.0G/401.9G (1.74%)
    default    4332.24500001/5498.0 (78.80%)  9.6T/27.2T (35.20%)   5.8T/46.3T (12.58%)
    jade       122.0/236.0 (51.69%)           655.3G/1.0T (62.24%)  178.7G/2.9T (6.10%)
    lucy       389.948/570.0 (68.41%)         2.5T/4.4T (56.57%)    52.0G/9.1T (0.56%)
    spam       0.0/1.0 (0.00%)                0B/5.8G (0.00%)       0B/67.0G (0.00%)
    yelp-main  1950.0/2190.0 (89.04%)         3.4T/3.9T (85.74%)    260.0G/34.8T (0.73%)
Marathon Status: OK
  marathon apps: 791
  marathon tasks: 4107
  marathon deployments: 3
Chronos Status: OK
  Enabled chronos jobs: 312
  Jobs Queued: 35 (11.22%)

[EvanKrall]

I guess I'm more interested in paasta metastatus -vvv, which has become my default invocation of paasta metastatus.