Commit Graph

192 Commits

Author SHA1 Message Date
Michael Schurter c8d3e869c6 Unregister from Consul when waiting for restart 2017-04-19 12:42:48 -07:00
Michael Schurter 16ac08ac8c Remove stale comment 2017-04-19 12:42:47 -07:00
Michael Schurter caf317e3f2 Use a DriverAbility to expose Exec functionality 2017-04-19 12:42:47 -07:00
Michael Schurter 0a59982936 Move removal from Consul into TaskRunner cleanup 2017-04-19 12:42:47 -07:00
Michael Schurter 745ad9521f Move ScriptExecutor to driver 2017-04-19 12:42:47 -07:00
Michael Schurter 244251490a Add UpdateTask method instead of Remove/Add 2017-04-19 12:42:47 -07:00
Michael Schurter e204a287ed Refactor Consul Syncer into new ServiceClient
Fixes #2478 #2474 #1995 #2294

The new client only handles agent and task service advertisement. Server
discovery is mostly unchanged.

The Nomad client agent now handles all Consul operations instead of the
executor handling task related operations. When upgrading from an
earlier version of Nomad existing executors will be told to deregister
from Consul so that the Nomad agent can re-register the task's services
and checks.

Drivers - other than qemu - now support an Exec method for executing
abritrary commands in a task's environment. This is used to implement
script checks.

Interfaces are used extensively to avoid interacting with Consul in
tests that don't assert any Consul related behavior.
2017-04-19 12:42:47 -07:00
Alex Dadgar 6bee23047a Fix variable capture and add tests
This PR fixes token revocation and adds tests to make sure it is
working. The 0.5.6 RC1's token revocation does not work becasue the
token's value is captured at the instantiation of the deferred
stoprenewal statement rather than its exectution.
2017-03-29 13:17:50 -07:00
Michael Schurter ae3810052d Merge pull request #2482 from hashicorp/f-2289-better-artifact-err
Improve artifact download error message
2017-03-28 12:48:22 -07:00
Alex Dadgar 1e95ae7e6a Merge pull request #2495 from hashicorp/b-vault-stop-renew
Stop Vault token renew on task exit
2017-03-28 11:14:18 -07:00
Alex Dadgar d1645f47b1 Stop Vault token renew on task exit
This PR fixes an oversight in which the client would attempt to renew a
token even after the task exits.

Fixes https://github.com/hashicorp/nomad/issues/2475
2017-03-28 10:53:15 -07:00
Michael Schurter 507862ade3 Add WrapRecoverable helper 2017-03-27 15:37:15 -07:00
Alex Dadgar 4ecebe7d8c Proper reference counting through task restarts
This PR fixes an issue in which the reference count on a Docker image
would become inflated through task restarts.
2017-03-25 17:05:53 -07:00
Michael Schurter 0e6c564406 Improve artifact download error message
Fixes #2289

Unfortunately took more RecoverableError hijinx than I would have liked.
There might be a better way.
2017-03-24 15:26:05 -07:00
Alex Dadgar 7e6c08191d Fix vet 2017-03-24 12:24:47 -07:00
Alex Dadgar c3551c761e Fix panic when restarting non-running task
This PR fixes an issue that is hit when running templates with restart
mode in which the client could panic when the handle is not running.

Fixes https://github.com/hashicorp/nomad/issues/2479
2017-03-24 12:04:22 -07:00
Alex Dadgar b5e53652aa Handle git ssh artifacts
This PR adds handling for downloading git artifacts using ssh with the
format git@github.com:hashicorp/go-getter.git

Fixes https://github.com/hashicorp/nomad/issues/2430
2017-03-11 15:12:41 -08:00
Alex Dadgar 3fb285f7d3 Fix TestAllocRunner_SaveRestoreState 2017-03-02 20:45:46 -08:00
Alex Dadgar 5cd43e837a Merge branch 'master' into b-remount 2017-03-02 19:23:13 -08:00
Michael Schurter e03f64ea6a Safely ensure {dev,proc,alloc} are mounted
If they're unmounted by a reboot they'll be properly remounted.
2017-03-02 13:21:34 -08:00
Alex Dadgar af4e400b36 Update go-getter and add support for git and hg
Fixes https://github.com/hashicorp/nomad/issues/2042
2017-03-01 14:46:04 -08:00
Alex Dadgar fa853c9696 Fix two issues during client restore state
This PR fixes two issues:

1) A close of a nil stopCollection channel when restoring and prestart
fails. The failure will cause the killCh to be triggered which will
close collection before it has been initialized.

2) Fixes a deadlock in which the handleWaitCh is never triggered since
it is not initialized when there is an error in prestart and the killCh
is triggered.

Both fixes are by maintaining the loop invariant that the two channels
are valid after there is a handle.
2017-02-28 10:29:12 -08:00
Alex Dadgar cef7882827 Fix tests and docs 2017-02-22 18:28:07 -08:00
Diptanu Choudhury 98921575af Adding a task event for setup 2017-02-22 18:28:07 -08:00
Michael Schurter 51e4fe9915 Use getters & setters with nil guards 2017-02-09 17:44:58 -08:00
Michael Schurter 37e7e7a3e5 Fix upgrade path for created resources
This *might* be a fix for #2295 -- I've been unable to reproduce the
bug. However, this guard seems wise regardless. I should never be
overwriting an intialized created resources with a nil.
2017-02-09 13:54:33 -08:00
Michael Schurter aef3c2e380 Handle createdResourcs=nil
Combined with b522c472fdf this fixes #2256

Without these two commits in place upgrades to 0.5.3 panics.
2017-01-31 10:51:32 -08:00
Alex Dadgar 8196a58c4c Rename dispatch_input to dispatch_payload 2017-01-25 21:27:44 -08:00
Michael Schurter 054ee8df59 Fix index we get allocs by 2017-01-20 16:30:40 -08:00
Michael Schurter c5f222e4a6 Update created resources before exiting cleanup 2017-01-19 16:48:23 -08:00
Michael Schurter a93d43a9cf Exit early when cleanup succeeds 2017-01-19 15:07:01 -08:00
Michael Schurter 85f68aa00c Fix incorrect lock usage 2017-01-19 11:39:18 -08:00
Michael Schurter a3a3656dbb Switch to use recoverable errors from Cleanup
TaskRunner handles retrying but Cleanup handles all of CreatedResources.
2017-01-13 16:46:08 -08:00
Michael Schurter dc68aa1a5a Return errors from cleanup and let TaskRunner retry 2017-01-12 17:21:54 -08:00
Michael Schurter 4d081490e6 Add Cleanup method to Driver interface
Cleanup can be used for cleaning up resources created by drivers to run
a task. Initially the Docker driver is the only user (to remove
downloaded images).
2017-01-11 17:23:33 -08:00
Alex Dadgar 4127a3ea9d Merge pull request #2173 from hashicorp/b-stats
Don't retrieve Driver Stats if unsupported
2017-01-10 13:32:03 -08:00
Alex Dadgar 2be221d664 Don't retrieve Driver Stats if unsupported
This PR makes us only try to collect stats once if the Driver doesn't
support collecting stats.

Fixes https://github.com/hashicorp/nomad/issues/1986
2017-01-09 13:47:06 -08:00
Alex Dadgar 26e2c5bb74 Merge pull request #2164 from hashicorp/b-dispatch
Create Task directory structure in the Run method
2017-01-09 11:24:46 -08:00
Alex Dadgar 2a5fd85e3b Move to Run() 2017-01-08 13:55:12 -08:00
Alex Dadgar 2affef2972 Create task directory during Prestart() 2017-01-08 13:55:12 -08:00
Alex Dadgar 4ffd9a69e5 Send Driver events to servers immediately
This PR causes driver events to be sent to the server immediately rather
than waiting for Prestart() to finish.
2017-01-08 13:54:43 -08:00
Michael Schurter 3ea09ba16a Move chroot building into TaskRunner
* Refactor AllocDir to have a TaskDir struct per task.
* Drivers expose filesystem isolation preference
* Fix lxc mounting of `secrets/`
2017-01-05 16:31:49 -08:00
Alex Dadgar 8d5f0fea69 Merge pull request #2128 from hashicorp/f-dispatch
Nomad Constructor Jobs and Dispatch
2017-01-06 05:22:49 +08:00
Michael Schurter ea92cd102a Append host env vars on every task env 2016-12-20 12:24:24 -08:00
Michael Schurter 2aa235f8f2 Rename InitializationMessage to DriverMessage 2016-12-20 11:51:09 -08:00
Alex Dadgar 159c819e08 Client writes payload to disk 2016-12-16 15:11:56 -08:00
Michael Schurter 770ed703d0 Add Driver.Prestart method
The Driver.Prestart method currently does very little but lays the
foundation for where lifecycle plugins can interleave execution _after_
task environment setup but _before_ the task starts.

Currently Prestart does two things:

* Any driver specific task environment building
* Download Docker images

This change also attaches a TaskEvent emitter to Drivers, so they can
emit events during task initialization.
2016-12-02 11:03:48 -08:00
Alex Dadgar 960424f086 Merge pull request #1941 from hashicorp/b-complete-transistion
Task state "dead" is terminal
2016-11-04 17:16:10 -07:00
Alex Dadgar e6465e138b More precise marking of dead 2016-11-04 17:11:07 -07:00
Alex Dadgar 0fb7742c3c Task state "dead" is terminal 2016-11-04 16:57:24 -07:00