7b3adc6cc4
client: refactor AllocRunner for testing
2015-08-30 16:35:04 -07:00
8c25d1fca1
client: testing task runner
2015-08-29 20:07:24 -07:00
db33f76a61
client: remove TaskRunner dependence on AllocRunner
2015-08-29 19:42:35 -07:00
a2d8c6c3fe
client: working on state restore
2015-08-29 19:14:47 -07:00
b6332b435d
client: first pass at save/restore of state
2015-08-29 18:16:49 -07:00
50c677a9bb
client: adding state save helpers
2015-08-29 18:03:00 -07:00
020ba19e5a
mock: make valid mock alloc
2015-08-29 16:24:01 -07:00
c99fe72f54
driver/exec: basic implementation
2015-08-29 16:20:07 -07:00
83dd313b5c
client: working on runners
2015-08-29 15:46:10 -07:00
e57ee3fbfe
client: test watching allocations
2015-08-29 14:33:30 -07:00
4d6238ebb8
client: test updating alloc status
2015-08-29 14:22:24 -07:00
2e4809d2fc
client: test heartbeats
2015-08-29 14:15:34 -07:00
5155f2c8eb
nomad: parameterize heartbeat configuration
2015-08-29 14:14:19 -07:00
25c7aa7e47
Merge pull request #12 from hashicorp/b-kernel.name
...
Update references to "os" to use "kernel.name"
2015-08-28 14:19:40 -07:00
ae1142ad33
cleanups
2015-08-28 13:09:21 -05:00
dc217da84a
spike on AWS ENV fingerprint
2015-08-28 11:33:41 -05:00
96cb220ff4
Update references to "os" to use "kernel.name"
...
This brings test code and mocks up to date with the fingerprinter. This was a slightly larger change than I anticipated, but I think it's good for two reasons:
1. More semanitcally correct. `os.name` is something like "Windows 10 Pro" or "Ubuntu", while `kernel.name` is "windows" or "linux". `os.version` and `kernel.version` match these semantics.
2. `kernel.name` is much easier to grep for than `os`, which is helpful because oracle can't help us with strings.
2015-08-28 01:30:47 -07:00
5a670e9380
Merge pull request #11 from hashicorp/b-windows
...
Windows bug fixes
2015-08-27 18:21:00 -07:00
9b8fc07bbe
Drop os fingerprint since this is redundant with host
2015-08-27 17:37:56 -07:00
94235e3448
Change 'os' to 'kernel.name'; add 'kernel.version' for *nix*
2015-08-27 17:37:37 -07:00
ced2f647c6
Add guard to storage fingerprint so CWD will be used if AllocDir is not specified -- fixes windows tests
2015-08-27 16:03:09 -07:00
58c7496975
Merge pull request #9 from hashicorp/b-storage-fingerprint
...
Storage fixes
2015-08-27 14:38:38 -07:00
d0a091c952
Add the storage fingerprinter to the list of built-ins
2015-08-27 14:30:41 -07:00
0437a02fd7
Change column constraint so this is compatible with darwin and linux
2015-08-27 14:30:20 -07:00
c6cdf014fd
Merge pull request #7 from hashicorp/memory-fingerprint
...
Add a memory fingerprint
2015-08-27 16:11:44 -05:00
965c9e176d
Merge pull request #8 from hashicorp/b-cpu-accumulation
...
Get average frequency of all CPUs so we can do average frequency * cores for total compute
2015-08-27 14:09:29 -07:00
b223278b6f
convert to MB for MemoryMB, and update test
2015-08-27 16:01:47 -05:00
6123ba8b83
Get average frequency of all CPUs so we can do average frequency * cores for total compute
2015-08-27 13:35:54 -07:00
759d3131f2
Merge remote-tracking branch 'upstream/master' into memory-fingerprint
...
* upstream/master:
Prevent a panic if the df tool output is not in the expected format
Change 'available' to 'free' because it's shorter; also initialize node.Resources.DiskMB as total disk instead of available disk
Changed logs to errors; added data to node.Resources.DiskMB
populate CPU in Node Resources
Fix arity in log call for TCP resolution failure
Add storage fingerprinter
2015-08-27 15:30:00 -05:00
e3348ad809
Merge pull request #6 from hashicorp/cpu-resources
...
populate CPU in Node Resources
2015-08-27 15:26:00 -05:00
83a792593e
Add a memory fingerprint, just grabs memory
2015-08-27 15:19:19 -05:00
78767f0ff4
Merge pull request #4 from hashicorp/f-storage-fingerprint
...
Add storage fingerprinter
2015-08-27 12:43:18 -07:00
8e544e0969
Prevent a panic if the df tool output is not in the expected format
2015-08-27 12:42:58 -07:00
66b518cd6c
Change 'available' to 'free' because it's shorter; also initialize node.Resources.DiskMB as total disk instead of available disk
2015-08-27 12:37:05 -07:00
d17c6128c8
Changed logs to errors; added data to node.Resources.DiskMB
2015-08-27 12:23:17 -07:00
58be8ba139
populate CPU in Node Resources
2015-08-27 14:15:56 -05:00
35196926d3
Merge pull request #5 from hashicorp/cpu-fingerprinting
...
Add cpu.frequency, cpu.totalcompute
2015-08-27 11:13:12 -05:00
cacb8074d3
Add cpu.frequency, cpu.totalcompute
2015-08-27 09:19:53 -05:00
cd5cd7efc2
Fix arity in log call for TCP resolution failure
2015-08-26 17:17:08 -07:00
edea3700a4
Add storage fingerprinter
...
Breaking change: Added ID() to Fingerprint interface. This allows us to assign
each fingerprint implementation a unique ID, identify which fingerprinters have
been run, and also self-identify in log messages.
- Added storage fingerprinter for Windows and *nix
- Added storage tests under storage_test.go
- Added test helper functions under fingerprint/fingerprint_test.go
- Added ID() to existing finterprinters and drivers
- Added Fingerprint.ID() to log messages via log.SetPrefix()
2015-08-26 17:17:08 -07:00
d64b749246
Merge pull request #3 from hashicorp/cpu-fingerprinting
...
Fingerprint CPU model name, count of cores
2015-08-26 17:01:08 -05:00
9a92383b69
refactor into multiple fingerprints
2015-08-26 16:29:18 -05:00
8ab31da8b0
update fingerprint with more cpu, some host info
2015-08-26 15:20:07 -05:00
d2318f71f1
Fingerprint CPU model name, count of cores
2015-08-26 11:32:30 -05:00
c7672b41ec
nomad: Adding Client.UpdateAlloc endpoint
2015-08-25 18:12:51 -07:00
a2010b9f2a
nomad: FSM support for AllocClientUpdateRequestType
2015-08-25 18:00:14 -07:00
922eb37484
nomad: adding UpdateAllocFromClient for client based updates
2015-08-25 17:54:45 -07:00
93a5b50483
Merge pull request #2 from hashicorp/f-alloc-status
...
Split the Alloc DesiredStatus from ClientStatus
2015-08-25 17:39:44 -07:00
2ea99f211a
nomad: updating for new alloc representation
2015-08-25 17:36:52 -07:00
2ee6947844
scheduler: updating for new APIs
2015-08-25 17:06:06 -07:00
Armon Dadgar