luxolus/open-nomad
2
0
Fork 0
Code Issues 1 Pull requests Packages Releases Wiki Activity
15476 commits 1 branch 0 tags 311 MiB
Commit graph

641 commits

Author SHA1 Message Date
Preetha Appan b46728a88b
Make spread weight a pointer with default value if unset 2019-01-11 10:31:21 -06:00
Michael Schurter 8a6b1acaa6 drain: fix node drain monitoring 
The whole approach to monitoring drains has ordering issues and lacks
state to output useful error messages.

AFAICT to get the tests passing reliably I needed to change the behavior
of monitoring.

Parts of these tests are skipped in CI, and they should be rewritten as
e2e tests.
 2019-01-08 09:35:16 -08:00
Nick Ethier a96afb6c91
fix tests that fail as a result of async client startup 2018-12-20 00:53:44 -05:00
Mahmood Ali d497729826
Merge pull request #4978 from hashicorp/f-device-tweaks 
Display device attributes in `nomad node status -verbose`
 2018-12-12 19:45:07 -05:00
Mahmood Ali fa9b9028a5 Use max 3 precision in displaying floats 
When formating floats in `nomad node status`, use a maximum precision of
3.
 2018-12-10 12:18:24 -05:00
Mahmood Ali 14668f48d1 device attributes in nomad node status -verbose 
This reports device attributes like the following:

```
$ nomad node status -self -verbose
ID          = f7adb958-29e1-2a5a-2303-9d61ffaab33a
Name        = mars.local
Class       = <none>
DC          = dc1
Drain       = false
Eligibility = eligible
Status      = ready
Uptime      = 12h40m13s

Drivers
Driver       Detected  Healthy  Message                               Time
docker       true      true     healthy                               2018-12-10T11:47:19-05:00
...

Attributes
cpu.arch                      = amd64
cpu.frequency                 = 2200
cpu.modelname                 = Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz
cpu.numcores                  = 12
...

Device Group Attributes
Device Group = nomad/file/mock
block_device = sda1
filesystem   = ext4
size         = 63.2 GB

Meta
```
 2018-12-10 12:18:24 -05:00
Alex Dadgar 1e3c3cb287 Deprecate IOPS 
IOPS have been modelled as a resource since Nomad 0.1 but has never
actually been detected and there is no plan in the short term to add
detection. This is because IOPS is a bit simplistic of a unit to define
the performance requirements from the underlying storage system. In its
current state it adds unnecessary confusion and can be removed without
impacting any users. This PR leaves IOPS defined at the jobspec parsing
level and in the api/ resources since these are the two public uses of
the field. These should be considered deprecated and only exist to allow
users to stop using them during the Nomad 0.9.x release. In the future,
there should be no expectation that the field will exist.
 2018-12-06 15:09:26 -08:00
Mahmood Ali f72e599ee7 Populate alloc stats API with device stats 
This change makes few compromises:

* Looks up the devices associated with tasks at look up time.  Given
that `nomad alloc status` is called rarely generally (compared to stats
telemetry and general job reporting), it seems fine.  However, the
lookup overhead grows bounded by number of `tasks x total-host-devices`,
which can be significant.

* `client.Client` performs the task devices->statistics lookup.  It
passes self to alloc/task runners so they can look up the device statistics
allocated to them.
  * Currently alloc/task runners are responsible for constructing the
entire RPC response for stats
  * The alternatives for making task runners device statistics aware
don't seem appealing (e.g. having task runners contain reference to hostStats)

* On the alloc aggregation resource usage, I did a naive merging of task device statistics.
  * Personally, I question the value of such aggregation, compared to
costs of struct duplication and bloating the response - but opted to be
consistent in the API.
  * With naive concatination, device instances from a single device group used by separate tasks in the alloc, would be aggregated in two separate device group statistics.
 2018-11-16 10:26:32 -05:00
Mahmood Ali 93e8fc53f9 device stats summary in node status 
Sample output with a mock device:

```
Host Resource Utilization
CPU             Memory          Disk
2651/26400 MHz  9.6 GiB/16 GiB  98 GiB/234 GiB

Device Resource Utilization
nomad/file/mock[README.md]    511 bytes
nomad/file/mock[e2e.go]       239 bytes
nomad/file/mock[e2e_test.go]  128 bytes

Allocations
No allocations placed
```
 2018-11-14 22:13:23 -05:00
Mahmood Ali 7d3e34fab5 Add NodeResource Device types in api package 2018-11-14 14:42:36 -05:00
Mahmood Ali 63acda956c Add Client Device Stats structs in api package 2018-11-14 14:41:19 -05:00
Preetha Appan fd0ba320da
change path to v1/scheduler/configuration 2018-11-12 15:57:45 -06:00
Preetha Appan 75662b50d1
Use response object/querymeta/writemeta in scheduler config API 2018-11-10 10:31:10 -06:00
Preetha Appan 5f0a9d2cfd
Show preemption output in plan CLI 2018-11-08 09:48:43 -06:00
Preetha Appan 57fe5050f0
more minor review feedback 2018-11-01 17:05:17 -05:00
Preetha Appan 8f7eb61823
Introduce a response object for scheduler configuration 2018-10-30 11:06:32 -05:00
Preetha Appan 1415032c13
More review comments 2018-10-30 11:06:32 -05:00
Preetha Appan c1c1c230e4
Make preemption config a struct to allow for enabling based on scheduler type 2018-10-30 11:06:32 -05:00
Preetha Appan bd34cbb1f7
Support for new scheduler config API, first use case is to disable preemption 2018-10-30 11:06:32 -05:00
Preetha Appan 9257387a69
Add number of evictions to DesiredUpdates struct to use in CLI/API 2018-10-30 11:06:32 -05:00
Preetha Appan 5b3bfb63eb
structs and API changes to plan and alloc structs needed for preemption 2018-10-30 11:06:32 -05:00
Alex Dadgar c176a83388 rename api TotalShares -> CpuShares 2018-10-16 17:25:55 -07:00
Alex Dadgar a78cefec18 use int64 2018-10-16 15:34:32 -07:00
Preetha Appan 7c0d8c646c
Change CPU/Disk/MemoryMB to int everywhere in new resource structs 2018-10-16 16:21:42 -05:00
Alex Dadgar 5a07f9f96e parse affinities and constraints on devices 2018-10-11 14:05:19 -07:00
Alex Dadgar 87cacb427f parse devices 2018-10-08 16:09:41 -07:00
Alex Dadgar 6b08b9d6b6 Define device request structs 2018-10-08 15:38:03 -07:00
Alex Dadgar bac5cb1e8b Scheduler uses allocated resources 2018-10-02 17:08:25 -07:00
Alex Dadgar 5c8697667e Node reserved resources 2018-09-29 18:44:55 -07:00
Alex Dadgar 3183153315 Node resources on client 2018-09-29 17:23:41 -07:00
Alex Dadgar cc92cd92cd
Merge pull request #4642 from hashicorp/b-vet 
Fix vet errors and use newer go version in travis
 2018-09-04 17:04:02 -07:00
Alex Dadgar c6576ddac1 Fix make check errors 2018-09-04 16:03:52 -07:00
Preetha Appan 751c0eb5a5
code review feedback 2018-09-04 16:10:11 -05:00
Preetha Appan 9bc0962527
Track top k nodes by norm score rather than top k nodes per scorer 2018-09-04 16:10:11 -05:00
Preetha Appan 6ed527c636
Use heap to store top K scoring nodes. 
Scoring metadata is now aggregated by scorer type to make it easier
to parse when reading it in the CLI.
 2018-09-04 16:10:11 -05:00
Preetha Appan 659cfa3f64
Parsing and API layer for spread stanza 2018-09-04 16:10:11 -05:00
Preetha Appan 1c0b123777
Fix test 2018-09-04 16:10:11 -05:00
Preetha Appan c407e3626f
More review comments 2018-09-04 16:10:11 -05:00
Preetha Appan 4b3b618e4a
Remove unused field 2018-09-04 16:10:11 -05:00
Preetha Appan 9f0caa9c3d
Affinity parsing, api and structs 2018-09-04 16:10:11 -05:00
Nick Ethier 41e010cdc2
nomad: add 'Dispatch' field to Job 
New -bash: Dispatch: command not found field is used to denote if the Job is a child dispatched job of
a parameterized job.
 2018-06-11 11:59:03 -04:00
Alex Dadgar 72effb8632 code review 2018-06-06 14:52:26 -07:00
Alex Dadgar f4fccd7ed2 Monitoring non-draining node exits 2018-06-05 17:58:44 -07:00
Alex Dadgar 7f25fcc1bd
Merge pull request #4354 from hashicorp/b-job-modify 
Deployment adds JobSpecModifyIndex
 2018-05-31 17:57:38 +00:00
Alex Dadgar f2b2e0482b code review fixes 2018-05-31 10:57:08 -07:00
Preetha Appan 4f835790d7
Set node eligibility to true when old client calls disable 2018-05-30 16:54:07 -05:00
Alex Dadgar 195e19827b Deployment adds JobSpecModifyIndex 
Deployment tracks the Job.JobModifyIndex so that PUTS against /v1/jobs
can be more easily coorelated with the created deployment.

Fixes https://github.com/hashicorp/nomad/issues/4301
 2018-05-30 11:33:56 -07:00
Nick Ethier 4b64db3a0f
api: emit different monitor message if node's drain strategy is never set 2018-05-24 06:39:09 -04:00
Preetha 159888a856
Merge pull request #4274 from hashicorp/f-force-rescheduling 
Add CLI and API support for forcing rescheduling of failed allocs
 2018-05-21 16:24:22 -07:00
Preetha Appan 4a400f045b
Fix docs and method documentation in API 2018-05-21 17:20:59 -05:00
Preetha Appan 3a8040e36f
Add new method EvaluateWithOptions to avoid breaking go API client 2018-05-11 14:18:53 -05:00
Preetha Appan b12df3c64b
Added CLI for evaluating job given ID, and modified client API for evaluate to take a request payload 2018-05-09 15:04:27 -05:00
Chelsea Holland Komlo d51611040f Add driver health information to node list stub 2018-05-09 11:21:54 -04:00
Alex Dadgar f4af30fbb5
Canary tags structs 2018-05-07 14:50:01 -05:00
Alex Dadgar f95ab4ade8
Mark canaries on creation, and unmark on promotion 2018-05-07 14:50:01 -05:00
Alex Dadgar 224b3092ae
change default to 10m and docs 2018-05-07 14:50:01 -05:00
Alex Dadgar 8a81038cdb
Set Reschedule from deployment watcher 2018-05-07 14:50:01 -05:00
Alex Dadgar e5caaf3358
Small test fix 2018-05-07 14:50:01 -05:00
Alex Dadgar 1336002255
Progress deadline in deployment state 2018-05-07 14:50:01 -05:00
Alex Dadgar ee50789c22
Initial implementation 2018-05-07 14:50:01 -05:00
Michael Schurter 0d534d30d6
Merge pull request #4251 from hashicorp/f-grpc-checks 
Support Consul gRPC Health Checks
 2018-05-04 14:55:16 -07:00
Michael Schurter 70b02875b7
Merge pull request #4234 from hashicorp/b-4159 
Fix race in StreamFramer and truncation in api/AllocFS.Logs
 2018-05-04 14:24:07 -07:00
Michael Schurter f6a4713141 consul: make grpc checks more like http checks 2018-05-04 11:08:11 -07:00
Michael Schurter 382caec1e1 consul: initial grpc implementation 
Needs to be more like http.
 2018-05-04 11:08:11 -07:00
Michael Schurter 526af6a246 framer: fix early exit/truncation in framer 2018-05-02 10:46:16 -07:00
Michael Schurter 949938534b api: never return EOF from Logs error chan 
Closing the frames chan is the only race-free way to signal to receivers
that all frames have been sent and no errors have occurred.

If EOF is sent on error chan receivers may not receive the last frame
(or frames since the chan is buffered) before receiving the error.

Closing frames is the idiomatic way of signaling there is no more data
to be read from a chan.
 2018-05-02 10:46:16 -07:00
Michael Schurter 6f0e2e808b tests: test logs from client<->api package 2018-05-02 10:46:16 -07:00
Preetha Appan 274bed1892
Add RescheduleTracker to allocs list stub struct 2018-05-01 14:53:47 -05:00
Alex Dadgar 15ad3f94af Fix command line 2018-04-26 15:46:22 -07:00
James Rasell b7c2ce2991
Update node-drain logging message to clearer for operators. 
This change updates the console log message when performing a node
drain and particulary when a node has marked all allocs for
migration. Previously it logged 'drain complete' which was a little
confusing to operators as the node is not drained at this point.

Closes #4183
 2018-04-24 07:50:01 +01:00
Nick Ethier 1f99c3a0f7
minor code review fixes to api/jobs 2018-04-17 10:18:36 -04:00
Nick Ethier 03a89060cf
api: add test for canonicalized jobs/parse 2018-04-16 19:21:09 -04:00
Nick Ethier de4176606d
command/agent: add Canonicalize option to parse args 2018-04-16 19:21:09 -04:00
Nick Ethier f2db03e56c
command/agent: add /v1/jobs/parse endpoint 
The parse endpoint accepts a hcl jobspec body within a json object
and returns the parsed json object for the job. This allows users to
register jobs with the nomad json api without specifically needing
a nomad binary to parse their hcl encoded jobspec file.
 2018-04-16 19:21:06 -04:00
Preetha bdc17ebf10
Merge pull request #4139 from hashicorp/b-reschedule-invalid-system-jobs 
Make system jobs fail validation if they contain a reschedule stanza
 2018-04-11 20:01:19 -05:00
Preetha Appan db549d1388
add canonicalize for reschedulepolicy to simplify validation logic 2018-04-11 18:47:27 -05:00
Preetha Appan ae1419826e
Always merge with default reschedule policy if its not nil 2018-04-11 15:26:01 -05:00
Preetha Appan a7b7b662ed
Make system jobs fail validation if they contain a reschedule stanza 2018-04-11 14:56:20 -05:00
Charlie Voiselle ba88f00ccb Changed "til" to "until" 
Should be "till" or "until"; chose "until" because it is unambiguous as to meaning.
 2018-04-11 12:36:28 -05:00
Alex Dadgar 3c51d5a5ea Fix ineffectual assignment 2018-04-05 11:29:39 -07:00
Alex Dadgar f1d3b47499 Don't assume the read index won't be zero if no jobs have been registered 2018-04-03 18:24:59 -07:00
Michael Schurter 7046db8818 cli: remove unreachable drain message 2018-03-30 14:15:12 -07:00
Michael Schurter f912cd4272 cli: log if a node goes down during draining 2018-03-30 14:02:42 -07:00
Michael Schurter 06874d8b3d drain: fix monitor node index handling 2018-03-30 12:43:53 -07:00
Michael Schurter 7199a2b960 cli: differentiate normal output vs info 2018-03-30 11:42:11 -07:00
Michael Schurter 0260bda046 cli: add color to drain output 2018-03-30 11:15:12 -07:00
Michael Schurter 6e10e0f84e drain: fix cli blocking when allocs already stopped 2018-03-30 10:18:14 -07:00
Michael Schurter ee3eddbac3 drain: block cli until all allocs stop 
Before the drain CLI would block until the node was marked as completing
drain operations. While technically correct, it could lead operators (or
more likely: scripts) to shutdown drained nodes before all of its
allocations had *actually* terminated.

This change makes the CLI block until all allocations have terminated
(unless ignoring system jobs).
 2018-03-29 10:56:09 -07:00
Alex Dadgar de4b3772f1 Create evals for system jobs when drain is unset 
This PR creates evals for system jobs when:

* Drain is unset and mark eligible is true
* Eligibility is restored to the node
 2018-03-27 15:53:24 -07:00
Chelsea Holland Komlo 003bc209b9 use time.Time for node events for compatibility 2018-03-27 15:43:57 -04:00
Preetha Appan 33e170c15d
s/linear/constant/g 2018-03-26 14:45:09 -05:00
Alex Dadgar d7f246efe1 Drain cli, api, http 2018-03-21 20:27:32 -07:00
Michael Schurter a854c7bdae docs: improve DrainRequest.MarkEligible comment 2018-03-21 16:55:22 -07:00
Michael Schurter a7ab75d853 test: index no longer guaranteed on job list 
Also switch to require and add t.Helper to appropriate funcs.
 2018-03-21 16:55:22 -07:00
Michael Schurter 1cc012966b api: fix tests to expect default migrate strategy 2018-03-21 16:51:45 -07:00
Michael Schurter 2832853bfa Add DesiredTransition.ShouldMigrate to api pkg 2018-03-21 16:51:45 -07:00
Alex Dadgar 92b636dd32 Fix deadline handling 2018-03-21 16:51:44 -07:00
Alex Dadgar 7b2bad8c5e Toggle Drain allows resetting eligibility 
This PR allows marking a node as eligible for scheduling while toggling
drain. By default the `nomad node drain -disable` commmand will mark it
as eligible but the drainer will maintain in-eligibility.
 2018-03-21 16:51:44 -07:00
Alex Dadgar 02019f216a Correct defaulting 2018-03-21 16:51:44 -07:00
Alex Dadgar a37329189a Improve DeadlineTime helper 2018-03-21 16:51:44 -07:00