Sandy Walsh 8a0f06ac79 Freshen up with latest from RackerLabs (and include tox.ini)

Added instance hours report

Initial version of report to calculate unit hours used
for nova instances

Breakdown by flavor, flavor class, account/billing types and by tenant.

Moved license so script has shebang as the first line
Add tenant info cache.
Refactor Instance hr report.
Added cache table for basic tenant info for reports.
Refactor instance_hours report to use table.
Improve performance of tenant info update.

use bulk sql operations to speed up the tenant info update,
as it's taking ~40s/1000 tenants to update on a decent machine.

Fix some tests broken by rebase. Fix unittests broken by
rebase. Also, renumber migration due to collision.

Add Apache license header to new files.

Fixed bug with fetching deployment information in
reconciler. Reverted old method for fetching
current usage's deployment and added new method to
fetch latest deployment information for
a request_id.

Made the field mismatch error message more readable
Refactored nova and glance verifier tests

the exists are updated with 201 send_status as part of stacktach down repair mechanism

Revert "Fixed bug with fetching deployment information in"

Revert "Adding host and deployment info to missing exists entries in the nova usage audit"

Revert "Added column headers for host and deployment in json reports"

Only log ERROR on last retry

fixed the wrong status name for sent_failed variable in audit report

fixing documentation for urls that are not available for glance

deprecating stacky urls (usage, deletes, exists) that are not
used anymore

Revert "Revert "Added column headers for host and deployment in json reports""

Revert "Revert "Adding host and deployment info to missing exists entries in the nova usage audit""

Revert "Revert "Fixed bug with fetching deployment information in""

Cell and compute info added for verification failures as well.
If that is not present(request_id is not populated for an
InstanceUsage entry), the cells display '-'

Add tox support for move to stackforge

Change-Id: Id94c2a7f1f9061e972e90c3f54e39c9dec11943b

2014-05-08 15:58:03 -03:00

17 KiB

Raw Permalink Blame History

The StackTach REST Interface

JSON Response Format

StackTach uses an tabular JSON response format to make it easier for Stacky to display generic results.

The JSON response format is as follows: :

[
  ['column header', 'column header', 'column header', ...],
  ['row 1, col 1', 'row 1, col 2', 'row 1, col 3', ...],
  ['row 2, col 1', 'row 2, col 2', 'row 2, col 3', ...],
  ['row 3, col 1', 'row 3, col 2', 'row 3, col 3', ...],
  ...
]

stacky/deployments

The list of all available deployments

Example request:

GET /stacky/deployments/ HTTP/1.1
Host: example.com
Accept: application/json

Example response:

HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ['#', 'Name'],
  [1, 'deployment name'],
  [2, 'deployment name'],
  ...
]

stacky/events

The distinct list of all event names

Example request:

GET /stacky/events/ HTTP/1.1
Host: example.com
Accept: application/json

Example response:

HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ['Event Name'],
  ["add_fixed_ip_to_instance"],
  ["attach_volume"],
  ["change_instance_metadata"],
  ["compute.instance.create.end"],
  ["compute.instance.create.error"],
  ["compute.instance.create.start"],
  ["compute.instance.create_ip.end"],
  ...
]

query service: nova or glance. default="nova"

stacky/hosts

The distinct list of all hosts sending notifications.

Example request:
GET /stacky/hosts/ HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ['Host Name'],
  ["compute-1"],
  ["compute-2"],
  ["scheduler-x"],
  ["api-88"],
  ...
query service

nova or glance. default="nova" ]

stacky/uuid

Retrieve all notifications for instances with a given UUID.

Example request:

GET /stacky/uuid/?uuid=77e0f192-00a2-4f14-ad56-7467897828ea  HTTP/1.1
Host: example.com
Accept: application/json

Example response:

HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ["#", "?", "When", "Deployment", "Event", "Host", "State",
   "State'", "Task"],
  [
      40065869,
      " ",
      "2014-01-14 15:39:22.574829",
      "region-1",
      "compute.instance.snapshot.start",
      "compute-99",
      "active",
      "",
      ""
  ],
  [
      40065879,
      " ",
      "2014-01-14 15:39:23.599298",
      "region-1",
      "compute.instance.update",
      "compute-99",
      "active",
      "active",
      "image_snapshot"
  ],
  ...
]

query uuid: UUID of desired instance.
query service: nova or glance. default="nova"

stacky/timings/uuid/

Retrieve all timings for a given instance. Timings are the time deltas between related .start and .end notifications. For example, the time difference between compute.instance.run_instance.start and compute.instance.run_instance.end. This url works only for nova.

The first column of the response will be

S if there is a .start event and no .end

E if there is a .end event and no .start

. if there was a .start and .end event

No time difference will be returned in the S or E cases.

Example request:
GET /stacky/timings/uuid/?uuid=77e0f192-00a2-4f14-ad56-7467897828ea  HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ["?", "Event", "Time (secs)"],
  [".", "compute.instance.create", "0d 00:00:55.50"],
  [".", "compute.instance.snapshot", "0d 00:14:11.71"],
  [".", "compute.instance.snapshot", "0d 00:17:31.33"],
  [".", "compute.instance.snapshot", "0d 00:16:48.88"]
  ...
]
query uuid

UUID of desired instance.

stacky/summary

Returns timing summary information for each event type collected. Only notifications with .start/.end pairs are considered. This url works only for nova.

This includes: :
* the number of events seen of each type (N)
* the Minimum time seen
* the Maximum time seen
* the Average time seen
Example request:
GET /stacky/summary/  HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ["Event", "N", "Min", "Max", "Avg"],
  ["compute.instance.create", 50,
      "0d 00:00:52.88", "0d 01:41:14.27", "0d 00:08:26"],
  ["compute.instance.create_ip", 50,
      "0d 00:00:06.80", "5d 20:16:47.08", "0d 03:47:17"],
  ...
]
query uuid

UUID of desired instance.

query limit

the number of timings to return.

query offset

offset into query result set to start from.

stacky/request

Returns all notifications related to a particular Request ID.

The ? column will be E if the event came from the .error queue. State and State' are the current state and the previous state, respectively. This url works only for nova.

Example request:
GET /stacky/request/?request_id=req-a7517402-6192-4d0a-85a1-e14051790d5a  HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ["#", "?", "When", "Deployment", "Event", "Host", "State",
   "State'", "Task'"
  ],
  [
      40368306,
      " ",
      "2014-01-15 15:39:34.130286",
      "region-1",
      "compute.instance.update",
      "api-1",
      "active",
      "active",
      null
  ],
  [
      40368308,
      " ",
      "2014-01-15 15:39:34.552434",
      "region-1",
      "compute.instance.update",
      "api-1",
      "active",
      null,
      null
  ],

  ...
]
query request_id

desired request ID

query when_min

unixtime to start search

query when_max

unixtime to end search

query limit

the number of timings to return.

query offset

offset into query result set to start from.

stacky/reports

Returns a list of all available reports.

The Start and End columns refer to the time span the report covers (in unixtime).

Example request:
GET /stacky/reports/ HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  ["Id", "Start", "End", "Created", "Name", "Version"],
  [
      5971,
      1389726000.0,
      1389729599.0,
      1389730212.9474499,
      "summary for region: all",
      4
  ],
  [
      5972,
      1389729600.0,
      1389733199.0,
      1389733809.979934,
      "summary for region: all",
      4
  ],

  ...
]
query created_from

unixtime to start search

query created_to

unixtime to end search

query limit

the number of timings to return.

query offset

offset into query result set to start from.

stacky/report/<report_id>

Returns a specific report.

The contents of the report varies by the specific report, but all are in row/column format with Row 0 being a special metadata row.

Row 0 of each report is a dictionary of metadata about the report. The actual row/columns of the report start at Row 1 onwards (where Row 1 is the Column headers and Rows 2+ are the details, as with other result sets)

Example request:

GET /stacky/report/1/ HTTP/1.1
Host: example.com
Accept: application/json

Example response:

HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  {
      "4xx failure count": 0,
      "4xx failure percentage": 0.0,
      "5xx failure count": 1,
      "5xx failure percentage": 0.018284904,
      "> 30 failure count": 13,
      "> 30 failure percentage": 1.13479794,
      "cells": [
          "c0001",
          "global",
          "c0003",
          "c0004",
          "c0011",
          "c0010",
          "a0001",
          "c0012",
          "b0002",
          "a0002"
      ],
      "end": 1389729599.0,
      "failure_grand_rate": 0.2445074415308293,
      "failure_grand_total": 14,
      "hours": 1,
      "pct": 0.014999999999999999,
      "percentile": 97,
      "region": null,
      "start": 1389726000.0,
      "state failure count": 0,
      "state failure percentage": 0.0,
      "total": 411
  },
  ["Operation", "Image", "OS Type", "Min", "Max", "Med", "97%", "Requests",
   "4xx", "% 4xx", "5xx", "% 5xx", "> 30", "% > 30", "state", "% state"],
  [
      "aux",
      "snap",
      "windows",
      "0s",
      "5s",
      "0s",
      "5s",
      6,
      0,
      0.0,
      0,
      0.0,
      0,
      0.0,
      0,
      0.0
  ],
  [
      "resize",
      "base",
      "linux",
      "1s",
      "5:44s",
      "1:05s",
      "3:44s",
      9,
      0,
      0.0,
      0,
      0.0,
      0,
      0.0,
      0,
      0.0
  ],

  ...
]

stacky/reports/search/

Returns reports that match the search criteria in descending order of id.

The contents of the report varies by the specific report, but all are in row/column format with Row 0 being a special metadata row. The actual row/columns of the report start at Row 1 onwards.

Example request:
GET /stacky/reports/search/ HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  [
      "Id",
      "Start",
      "End",
      "Created",
      "Name",
      "Version"
  ],

  [
      4253,
      "2013-11-21 00:00:00",
      "2013-11-22 00:00:00",
      "2013-11-22 01:44:55",
      "public outbound bandwidth",
      1
  ],
  [
      4252,
      "2014-01-18 00:00:00",
      "2013-11-22 00:00:00",
      "2013-11-22 01:44:55",
      "image events audit",
      1
  ],
  [
      4248,
      "2013-11-21 00:00:00",
      "2013-11-22 00:00:00",
      "2013-11-22 01:44:55",
      "Error detail report",
      1
  ],

  ...
]
query id

integer report id

query name

string report name(can include spaces)

query period_start

start of period, which the report pertains to, in the following format: YYYY-MM-DD HH:MM[:ss[.uuuuuu]][TZ]

query period_end

end of period, which the report pertains to, in the following format: YYYY-MM-DD HH:MM[:ss[.uuuuuu]][TZ]

query created

the day, when the report was created, in the following format: YYYY-MM-DD

stacky/show/<event_id>

Show the details on a specific notification.

The response of this operation is non-standard. It returns 3 rows:

The first row is the traditional row-column result set used by most commands.
The second row is a prettied, stringified version of the full JSON payload of the raw notification.
The third row is the UUID of the related instance, if any.

Example request:

GET /stacky/show/1234/  HTTP/1.1
Host: example.com
Accept: application/json

Example response:

HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  [
    ["Key", "Value"],
    ["#", 1234 ],
    ["When", "2014-01-15 20:39:44.277745"],
    ["Deployment", "region-1"],
    ["Category", "monitor.info"],
    ["Publisher", "compute-1"],
    ["State", "active"],
    ["Event", "compute.instance.update"],
    ["Service", "compute"],
    ["Host", "compute-1"],
    ["UUID", "8eba1a6d-43eb-1343-8d1a-5e596f5233b5"],
    ["Req ID", "req-1368539d-f645-4d96-842e-03b5c5c9dc8c"],
    ...
  ],
  "[\n  \"monitor.info\", \n  {\n    \"_context_request_id\": \"req-13685e9d-f645-4d96-842e-03b5c5c9dc8c\", \n    \"_context_quota_class\": null, \n    \"event_type\": \"compute.instance.update\", \n    \"_context_service_catalog\": [], \n    \"_context_auth_token\": \"d81a25d03bb340bb82b4b67d105cc42d\", \n    \"_context_user_id\": \"b83e2fac644c4215bc449fb4b5c9bbfa\", \n    \"payload\": {\n      \"state_description\": \"\", \n      \"availability_zone\": null, \n      \"terminated_at\": \"\", \n      \"ephemeral_gb\": 300, \n ...",
  "8eba1a6d-43eb-1343-8d1a-5e596f5233b5"
]

query service: nova or glance. default="nova"
query event_id: desired Event ID

stacky/watch/<deployment_id>

Get a real-time feed of notifications.

Once again, this is a non-standard response (not the typical row-column format). This call returns a tuple of information:

A list of column widths, to be used as a hint for formatting.

A list of events that meet the query criteria.

the db id of the event

the type of event (E for errors, . otherwise)

stringified date of the event

stringified time of the event

deployment name

the event name

the instance UUID, if available

The ending unixtime timestamp. The last time covered by this query (utcnow, essentially)

Example request:
GET /stacky/watch/14/  HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  [10, 1, 15, 20, 50, 36],
  [
  ... events ...
  ]
  "1389892207"
]
query service

nova or glance. default="nova"

query since

get all events since unixtime. Defaults to 2 seconds ago.

query event_name

only watch for event_name notifications. Defaults to all events.

stacky/search

Search for notifications.

Returns:

Event ID

E for errors, . otherwise

unixtime for when the event was generated

the deployment name

the event name

the host name

the instance UUID

the request ID

Example request:
GET /stacky/search/  HTTP/1.1
Host: example.com
Accept: application/json
Example response:
HTTP/1.1 200 OK
Vary: Accept
Content-Type: text/json

[
  [...event info as listed above...]
]
query service

nova or glance. default="nova"

query field

notification field to search on.

query value

notification values to find.

query when_min

unixtime to start search

query when_max

unixtime to end search

17 KiB Raw Permalink Blame History

The StackTach REST Interface

JSON Response Format

stacky/deployments

stacky/events

stacky/hosts

stacky/uuid

stacky/timings/uuid/

stacky/summary

stacky/request

stacky/reports

stacky/report/<report_id>

stacky/reports/search/

stacky/show/<event_id>

stacky/watch/<deployment_id>

stacky/search

17 KiB

Raw Permalink Blame History