Commit graph

13146 commits

Author SHA1 Message Date
Nick Ethier 5617f3615b driver/raw_exec: initial raw_exec implementation 2018-10-16 16:53:31 -07:00
Nick Ethier bcc5c4a8bd clientv2: base driver plugin (#4671)
Driver plugin framework to facilitate development of driver plugins.

Implementing plugins only need to implement the DriverPlugin interface.
The framework proxies this interface to the go-plugin GRPC interface generated
from the driver.proto spec.

A testing harness is provided to allow implementing drivers to test the full
lifecycle of the driver plugin. An example use:

func TestMyDriver(t *testing.T) {
    harness := NewDriverHarness(t, &MyDiverPlugin{})
    // The harness implements the DriverPlugin interface and can be used as such
    taskHandle, err := harness.StartTask(...)
}
2018-10-16 16:53:31 -07:00
Michael Schurter 62c1285afc tr: add comments and cleanup call signature
From review comments on #4649 left post-merge.
2018-10-16 16:53:31 -07:00
Nick Ethier 5dee1141d1 executor v2 (#4656)
* client/executor: refactor client to remove interpolation

* executor: POC libcontainer based executor

* vendor: use hashicorp libcontainer fork

* vendor: add libcontainer/nsenter dep

* executor: updated executor interface to simplify operations

* executor: implement logging pipe

* logmon: new logmon plugin to manage task logs

* driver/executor: use logmon for log management

* executor: fix tests and windows build

* executor: fix logging key names

* executor: fix test failures

* executor: add config field to toggle between using libcontainer and standard executors

* logmon: use discover utility to discover nomad executable

* executor: only call libcontainer-shim on main in linux

* logmon: use seperate path configs for stdout/stderr fifos

* executor: windows fixes

* executor: created reusable pid stats collection utility that can be used in an executor

* executor: update fifo.Open calls

* executor: fix build

* remove executor from docker driver

* executor: Shutdown func to kill and cleanup executor and its children

* executor: move linux specific universal executor funcs to seperate file

* move logmon initialization to a task runner hook

* client: doc fixes and renaming from code review


* taskrunner: use shared config struct for logmon fifo fields

* taskrunner: logmon only needs to be started once per task
2018-10-16 16:53:31 -07:00
Michael Schurter e6e2930a00 tr: implement stats collection hook
Tested except for the net/rpc specific error case which may need
changing in the gRPC world.
2018-10-16 16:53:31 -07:00
Michael Schurter 86bd329539 fix build errors post merges 2018-10-16 16:53:31 -07:00
Michael Schurter a977e22028 test: cleanup mock consul service client
Updated to hclog.

It exposed fields that required an unexported lock to access. Created a
getter methodn instead. Only old allocrunner currently used this
feature.
2018-10-16 16:53:31 -07:00
Michael Schurter 6f92b04226 health_hook: simplify locking; test thoroughly
Use doneCh like @dadgar suggested in the original PR.

Thoroughly test hook as concurrent Update calls make for a tricky
concurrency problem.
2018-10-16 16:53:30 -07:00
Alex Dadgar cebfead6bc add logger back 2018-10-16 16:53:30 -07:00
Nick Ethier 03422aa529 fifo: add new fifo package for named pipes (#4665)
* fifo: add new fifo package for named pipes
2018-10-16 16:53:30 -07:00
Alex Dadgar 8504505c0d client uses passed logger and fix fingerprinters 2018-10-16 16:53:30 -07:00
Nick Ethier 66ff12e5f7 Update runc/libcontainer and friends (#4655)
* vendor: bump libcontainer and docker to remove Sirupsen imports

* vendor: fix bad vendoring of archive package

* vendor: fix api changes to cgroups in executor

* vendor: fix docker api changes

* vendor: update github.com/Azure/go-ansiterm to use non capitalized logrus import
2018-10-16 16:53:30 -07:00
Michael Schurter 195b8127fb health_hook: fix panic and add tests
Still more testing to do, but I want to get this panic fixed ASAP.

All new tests pass with -race
2018-10-16 16:53:30 -07:00
Michael Schurter 64efc3d301 Emit events before long operations
Append when there's nothing blocking between appending and sending an
update to the server.
2018-10-16 16:53:30 -07:00
Michael Schurter a2b696c4cf Use a semaphore to block until watcher exits 2018-10-16 16:53:30 -07:00
Michael Schurter a73162c977 ar: use multierror in update hook loop
Make it match TaskRunner update hook behavior
2018-10-16 16:53:30 -07:00
Michael Schurter a7b427718c tr: refactor EmitEvents into Emit+Append
* UpdateState: set state, append event, persist, update servers
* EmitEvent: append event, persist, update servers
* AppendEvent: append event, persist

AppendEvent may not even have to persist, but for the sake of
correctness I'm going with that for now.
2018-10-16 16:53:30 -07:00
Michael Schurter 93f3ac9ed6 ar: create health setting shim for health watcher 2018-10-16 16:53:30 -07:00
Michael Schurter 4d5aaac6d2 fix detection of task transitioning to running 2018-10-16 16:53:30 -07:00
Michael Schurter 4136e59f79 arv2: implement alloc health watching
Also remove initial alloc from broadcaster as it just caused useless
extra processing.
2018-10-16 16:53:30 -07:00
Michael Schurter 5c5c6dc41b refactor ar hooks into their own files
minimize passed dependencies to ease testing
2018-10-16 16:53:30 -07:00
Michael Schurter 0bbf3a93ee make AllocBroadcaster easier to use
And test thoroughly.
2018-10-16 16:53:30 -07:00
Michael Schurter 9d1ea3b228 client: hclog-ify most of the client
Leaving fingerprinters in case that interface changes with plugins.
2018-10-16 16:53:30 -07:00
Michael Schurter e42154fc46 implement stopping, destroying, and disk migration
* Stopping an alloc is implemented via Updates but update hooks are
  *not* run.
* Destroying an alloc is a best effort cleanup.
* AllocRunner destroy hooks implemented.
* Disk migration and blocking on a previous allocation exiting moved to
  its own package to avoid cycles. Now only depends on alloc broadcaster
  instead of also using a waitch.
* AllocBroadcaster now only drops stale allocations and always keeps the
  latest version.
* Made AllocDir safe for concurrent use

Lots of internal contexts that are currently unused. Unsure if they
should be used or removed.
2018-10-16 16:53:30 -07:00
Michael Schurter 4236255686 lots of comment/log fixes 2018-10-16 16:53:30 -07:00
Michael Schurter 5749ede04e keep forgetting lxc 2018-10-16 16:53:30 -07:00
Michael Schurter 357641c364 persist alloc state on changes, not periodically
Allow alloc and task runners to persist their own state when something
changes instead of periodically syncing all state.
2018-10-16 16:53:30 -07:00
Michael Schurter 820af27171 wrap boltdb in a write deduplicator
Saves a tiny bit of cpu and some IO. Sadly doesn't prevent all IO on
duplicate writes as the transactions are still created and committed.

$ go test -bench=. -benchmem
goos: linux
goarch: amd64
pkg: github.com/hashicorp/nomad/helper/boltdd
BenchmarkWriteDeduplication_On-4             500           4059591 ns/op           23736 B/op         56 allocs/op
BenchmarkWriteDeduplication_Off-4            300           4115319 ns/op           25942 B/op         55 allocs/op
2018-10-16 16:53:30 -07:00
Michael Schurter 990228a6e2 wip wrap boltdb to get path information
finished but doesn't handle deleting deeply nested buckets
2018-10-16 16:53:30 -07:00
Michael Schurter a3fe0510d1 Move all encoding and put deduping into state db
Still WIP as it does not handle deletions.
2018-10-16 16:53:30 -07:00
Michael Schurter 533bc93b3a implement all boltdb interactions behind StateDB 2018-10-16 16:53:30 -07:00
Michael Schurter d890de036a tr: persist hook state whenever it changes 2018-10-16 16:53:30 -07:00
Michael Schurter fae5e89a0e artifacts: don't emit event when there's no artifacts 2018-10-16 16:53:30 -07:00
Michael Schurter 5383d20505 removing old restoration path before api change 2018-10-16 16:53:30 -07:00
Michael Schurter a5d3e3fb0a Implement alloc updates in arv2
Updates are applied asynchronously but sequentially
2018-10-16 16:53:30 -07:00
Michael Schurter 39b3f3a85b call handle.Network() instead of storing it 2018-10-16 16:53:30 -07:00
Michael Schurter 7132b67c1e Add Network method to Handle interface
Should probably be moved to an Inspect method in the Driver Plugin world
2018-10-16 16:53:30 -07:00
Michael Schurter a4b4d7b266 consul service hook
Deregistration works but difficult to test due to terminal updates not
being fully implemented in the new client/ar/tr.
2018-10-16 16:53:29 -07:00
Michael Schurter 5be982e674 restore vault client 2018-10-16 16:53:29 -07:00
Michael Schurter ce04915c9f log before killing tasks 2018-10-16 16:53:29 -07:00
Michael Schurter a2bf851805 no need to TaskStateUpdated to return an error
also updated comments
2018-10-16 16:53:29 -07:00
Alex Dadgar fd3bc1bd39 Update state with server 2018-10-16 16:53:29 -07:00
Alex Dadgar bc905cc61d Define and thread through state updating interface 2018-10-16 16:53:29 -07:00
Michael Schurter 9a63d6103d tr: add validate task hook 2018-10-16 16:53:29 -07:00
Michael Schurter 7f4ec50906 missed locking around c.allocs access 2018-10-16 16:53:29 -07:00
Alex Dadgar c93cfc89c0 wip 2018-10-16 16:53:29 -07:00
Alex Dadgar 7ddc0eb65c Fix deadlock 2018-10-16 16:53:29 -07:00
Alex Dadgar 3779077052 Remove SetState from interface 2018-10-16 16:53:29 -07:00
Alex Dadgar e1ba73b515 compile 2018-10-16 16:53:29 -07:00
Michael Schurter 6ebdf532ea wip split event emitting and state transitions 2018-10-16 16:53:29 -07:00