Commit Graph

1622 Commits

Author SHA1 Message Date
Michael Schurter 694b547a6b
Merge pull request #3682 from hashicorp/b-3681-always-set-driver-ip
Always advertise driver IP when in driver mode
2018-01-22 16:41:34 -08:00
Kyle Havlovitz 901c8791f4
Merge pull request #3670 from hashicorp/autopilot
Add Autopilot feature from Consul
2018-01-19 12:52:56 -08:00
Michael Schurter 8a0cf66822 Improve invalid port error message for services
Related to #3681

If a user specifies an invalid port *label* when using
address_mode=driver they'll get an error message about the label being
an invalid number which is very confusing.

I also added a bunch of testing around Service.AddressMode validation
since I was concerned by the linked issue that there were cases I was
missing. Unfortunately when address_mode=driver is used there's only so
much validation that can be done as structs/structs.go validation never
peeks into the driver config which would be needed to verify the port
labels/map.
2018-01-18 15:35:24 -08:00
Michael Schurter 447dc5bbd3 Fix test 2018-01-18 15:35:24 -08:00
Michael Schurter 583e17fad5 Always advertise driver IP when in driver mode
Fixes #3681

When in drive address mode Nomad should always advertise the driver's IP
in Consul even when no network exists. This matches the 0.6 behavior.

When in host address mode Nomad advertises the alloc's network's IP if
one exists. Otherwise it lets Consul determine the IP.

I also added some much needed logging around Docker's network discovery.
2018-01-18 15:35:24 -08:00
Kyle Havlovitz 8d41f4ad40 Formatting/test adjustments 2018-01-18 15:03:35 -08:00
Kyle Havlovitz 12ff22ea70 Merge branch 'master' into autopilot 2018-01-18 13:29:25 -08:00
Michael Schurter 9fcda9c5aa Make example service name make sense
Switch from global-redis-check for the example job's service name to
redis-cache. The former name is really confusing and someone finally
called us out on it:

https://groups.google.com/d/msg/nomad-tool/3RTh6CyYkWk/vEe_Sj7lAAAJ

Also specifically mention that the `service.name` parameter is what is
advertised in Consul.
2018-01-17 14:08:30 -08:00
Kyle Havlovitz bc385bcc93 Fix comments/text referring to consul 2018-01-17 00:20:13 -08:00
Michael Schurter 57eb128dcf
Merge pull request #3718 from hashicorp/b-3713-fix-check-restart
Fix service.check_restart stanza propagation
2018-01-16 16:39:42 -08:00
Kyle Havlovitz 7b980c42d8 Add raft remove by id endpoint/command 2018-01-16 13:35:32 -08:00
Alex Dadgar 54124a8478 Test listener uses freeport instead of static ports 2018-01-12 15:10:26 -08:00
Michael Schurter 9f179e9fab Fix HTTP code for permission denied errors
Fixes #3697

The existing code and test case only covered the leader behavior. When
querying against non-leaders the error has an "rpc error: " prefix.

To provide consistency in HTTP error response I also strip the "rpc
error: " prefix for 403 responses as they offer no beneficial additional
information (and in theory disclose a tiny bit of data to unauthorized
users, but it would be a pretty weird bit of data to use in a malicious
way).
2018-01-09 15:25:53 -08:00
Michael Schurter 7c282f174b Fix service.check_restart stanza propagation
There was a bug in jobspec parsing, a bug in CheckRestart merging, and a
bug in CheckRestart canonicalization. All are now tested.
2018-01-09 15:15:36 -08:00
Preetha 1712b03705
Merge branch 'master' into 0.8 2018-01-03 16:06:38 -06:00
Alex Dadgar bfc62ae41c bump version and remove generated structs 2017-12-19 17:10:52 -08:00
Alex Dadgar f0127afd93 generated files 2017-12-19 16:57:34 -08:00
Michael Schurter 714eb0b266 Services should not require a port
Fixes #3673
2017-12-19 15:50:23 -08:00
Kyle Havlovitz 1c07066064 Add autopilot functionality based on Consul's autopilot 2017-12-18 14:29:41 -08:00
Kyle Havlovitz b775fc7b33
Added support for v2 raft APIs and -raft-protocol option 2017-12-12 10:17:16 -06:00
Alex Dadgar d61ade8f02 remove generated structs 2017-12-11 17:51:41 -08:00
Alex Dadgar 8e63d545c4 generated assets 2017-12-11 17:30:37 -08:00
Michael Schurter b6152cf436
Merge pull request #3650 from hashicorp/b-lowercase-agent-service-id
Use the Service.Hash() method in agent service ids
2017-12-11 17:12:52 -08:00
Michael Schurter cdcefd0908 Use the Service.Hash() method in agent service ids
The allocID and taskName parameters are useless for agents, but it's
still nice to reuse the same hash method for agent and task services.
This brings in the lowercase mode for the agent hash as well.
2017-12-11 16:50:15 -08:00
Alex Dadgar 45f5f4c521
Update job_status.go 2017-12-11 16:14:03 -08:00
Alex Dadgar 9a8a89c3ab Node Status output includes new modify time 2017-12-11 16:08:25 -08:00
Alex Dadgar 189ff0dc22 Fix upgrade path of modify time 2017-12-11 15:58:24 -08:00
Michael Schurter 4f1002c1a8 Be more defensive in port checks 2017-12-08 12:27:57 -08:00
Michael Schurter d613e0aaf5 Move service hash logic to Service.Hash method 2017-12-08 12:03:43 -08:00
Michael Schurter b71edf846f Hash fields used in task service IDs
Fixes #3620

Previously we concatenated tags into task service IDs. This could break
deregistration of tag names that contained double //s like some Fabio
tags.

This change breaks service ID backward compatibility so on upgrade all
users services and checks will be removed and re-added with new IDs.

This change has the side effect of including all service fields in the
ID's hash, so we no longer have to track PortLabel and AddressMode
changes independently.
2017-12-08 12:03:43 -08:00
Michael Schurter 91282315d1 Prevent using port 0 with address_mode=driver 2017-12-08 12:03:43 -08:00
Michael Schurter 4b20441eef Validate port label for host address mode
Also skip getting an address for script checks which don't use them.

Fixed a weird invalid reserved port in a TaskRunner test helper as well
as a problem with our mock Alloc/Job. Hopefully the latter doesn't cause
other tests to fail, but we were referencing an invalid PortLabel and
just not catching it before.
2017-12-08 12:03:43 -08:00
Michael Schurter 4347026f83 Test Consul from TaskRunner thoroughly
Rely less on the mockConsulServiceClient because the real
consul.ServiceClient needs all the testing it can get!
2017-12-08 12:03:00 -08:00
Michael Schurter 4ae115dc59 Allow custom ports for services and checks
Fixes #3380

Adds address_mode to checks (but no auto) and allows services and checks
to set literal port numbers when using address_mode=driver.

This allows SDNs, overlays, etc to advertise internal and host addresses
as well as do checks against either.
2017-12-08 12:03:00 -08:00
Michael Schurter 1dd5b3822c
Merge pull request #3608 from hashicorp/b-3342-windows-log-leak
Fix bug in log framer only affecting Windows
2017-12-08 10:59:26 -08:00
Chelsea Holland Komlo 2ea8e43214 code review fixups 2017-12-06 16:37:47 -05:00
Chelsea Holland Komlo a010db084b fix up basic test
add conversion for KillSignal for api/struct representation of task
2017-12-06 14:36:45 -05:00
Preetha 4e6204777e
Merge pull request #3618 from hashicorp/b_fix_ns_inspect_test
Fix Test: Inspect returns json, not cli formatted output
2017-12-05 12:41:04 -06:00
Michael Schurter b66aa5b7f6
Merge pull request #3563 from hashicorp/b-snapshot-atomic
Atomic Snapshotting / Sticky Volume Migration
2017-12-05 09:16:33 -08:00
Charlie Voiselle f7c4156f95 Inspect returns json, not cli formatted output 2017-12-05 12:04:11 -05:00
Alex Dadgar 0bec137561
Merge pull request #3555 from PagerDuty/fix-loop-on-sigpipe
Do not emit logs on SIGPIPE since logging service could be unavailable
2017-12-04 14:11:05 -08:00
Alex Dadgar ab67a98c13 Emit hostname as a label 2017-12-04 10:42:31 -08:00
Jens Herrmann 5680fcccc2 Fix typos in metric names. #3610 2017-12-01 15:24:14 +01:00
Michael Schurter 2cbde16b9b Add check for Windows ECONNRESET 2017-11-30 21:30:20 -08:00
Michael Schurter 3e8e3aac70 Add defensive check to safeguard from future #3342s
I hate adding "this should never happen" checks, but causing a tight
loop that OOMs Nomad is just too easy in this code otherwise.
2017-11-30 20:37:13 -08:00
Michael Schurter 29d86eb348 Fix race in framer and improperly returned err
Fixes #3342

Two bugs were fixed:

* Closing the StreamFramer's exitCh before setting the error means other
  goroutines blocked on exitCh closing could see the error as nil. This
  was *not* observered.
* parseFramerError on Windows would fall through and return an
  improperly captured nil err variable. There's no need for
  parseFramerError to be a closure which fixes the confusion.
2017-11-30 17:42:53 -08:00
Michael Schurter 5e975bbd0f Add comment and normalize err check ordering
as per PR comments
2017-11-29 17:26:11 -08:00
Michael Schurter d996c3a231 Check for error file when receiving snapshots 2017-11-29 17:26:11 -08:00
Michael Schurter ca946679f6 Destroy partially migrated alloc dirs
Test that snapshot errors don't return a valid tar currently fails.
2017-11-29 17:26:11 -08:00
Charlie Voiselle c563aa7bd9 gofmt changes 2017-11-29 19:43:19 -05:00