Commit graph

4457 commits

Author SHA1 Message Date
Sean Chittenden fa7285cb5b
Don't spam the consul if Consul is not available.
Log once when Consul goes away, and log when Consul comes back.
2016-06-10 15:54:39 -04:00
Sean Chittenden a1a0c313de
Skip nil check for agent's consulSyncer is always not nil 2016-06-10 15:54:39 -04:00
Sean Chittenden a2109af862
Properly cover Syncer attributes with the registryLock.
trackedServices, delegateChecks, trackedChecks, and checkRunners
should all be covered.  This lock needs to be reasonably narrow and
can't use defer due to possible recursive locking concerns further
downstream from the call sites.
2016-06-10 15:54:39 -04:00
Sean Chittenden cc6e8792e0
Create a consulContext using a client's consul config.
This is wrong and should be the Agent's Consul Config.  This is a
step in the right direction, so committing to mark the necessary
future change.
2016-06-10 15:54:39 -04:00
Sean Chittenden f9b561b52f
On Syncer Shutdown, remove all services that match a Syncer's prefix. 2016-06-10 15:54:39 -04:00
Sean Chittenden 077203fe93
Update the structure of ConsulService to match reality.
ConsulService is the configuration for a Consul Service
2016-06-10 15:54:39 -04:00
Sean Chittenden 57e084e4df
Sync checks with Consul by comparing the AgentCheckReg w/ ConsulService
The source of truth is the local Nomad Agent.  Any checks are not local that
have a matching prefix are removed.  Changed checks are re-registered
and missing checks are re-added.
2016-06-10 15:54:39 -04:00
Sean Chittenden 197feae679
Sync services with Consul by comparing the AgentServiceReg w/ ConsulService
The source of truth is the local Nomad Agent.  Any services not local that
have a matching prefix are removed.  Changed services are re-registered
and missing services are re-added.
2016-06-10 15:54:39 -04:00
Sean Chittenden ed29946f5e
Populate the RPC Proxy's server list if heartbeat did not include a leader.
It's possible that a Nomad Client is heartbeating with a Nomad server that
has become issolated from the quorum of Nomad Servers.  When 3x the
heartbeatTTL has been exceeded, append the Consul server list to the primary
primary server list.  When the next RPCProxy rebalance occurs, there is a
chance one of the servers discovered from Consul will be in the majority.
When client reattaches to a Nomad Server in the majority, it will include
a heartbeat and will reset the TTLs *AND* will clear the primary server list
to include only values from the heartbeat.
2016-06-10 15:54:39 -04:00
Sean Chittenden 9a223936bb
Generate and sync Consul ServiceIDs consistently 2016-06-10 15:54:39 -04:00
Sean Chittenden bcad11cc2a
Rename runChecks to consulAvailable
Apologies in advance for the variable thrash, the fingerprinter is
no longer used to gate whether or not Consul is available any more.
2016-06-10 15:54:39 -04:00
Sean Chittenden efe00d2ccd
Update Syncer.Run() to call SyncServices(). 2016-06-10 15:54:39 -04:00
Sean Chittenden 83775c5c8b
Add "Service Groups" to the Syncer.
Now the right way to register services with the Syncer is to call
`SetServices(groupName, []*services)`.  This was required to allow
the Syncer to sync either the Client, Server, or Both using a
single Syncer.
2016-06-10 15:54:39 -04:00
Sean Chittenden c54539b860
Initialize Consul for the Nomad Agent in a more uniform way.
Decompose Client and Server registration into `setupClient()` and
`setupServer()`, respectively.
2016-06-10 15:54:39 -04:00
Sean Chittenden 95c9d1a63e
Per-comment, remove structs.Allocation's Services attribute.
Nuke PopulateServiceIDs() now that it's also no longer needed.
2016-06-10 15:54:39 -04:00
Sean Chittenden 06e5e615a8
Rename command/agent/consul/sync.go to syncer.go 2016-06-10 15:54:39 -04:00
Sean Chittenden d307ca2c28
Begin leveraging the Agent-level consul.Syncer 2016-06-10 15:54:39 -04:00
Sean Chittenden 3892a0433e
Move the start of the UniversalExecutor's consulSyncer to initialize once
This should be handled via a sync.Once primative, but I don't want to
unpack that atm.
2016-06-10 15:54:39 -04:00
Sean Chittenden 7956eb0c80
Rename structs.Task's Service attribute to ConsulService 2016-06-10 15:54:39 -04:00
Sean Chittenden 05a2b9528c
Remove Syncer.registerService()
This call is obsolete by a future commit that changes the canonical
source of truth to be consul.AgentServiceRegistration structs, which
means it is not necessary to construct AgentServiceRegistration
objects every time a registration is made, we just reuse the existing
object.
2016-06-10 15:54:39 -04:00
Sean Chittenden 8c813630e6
Move package client/consul/sync to command/agent/consul.
This has been done to allow the Server and Client to reuse the same
Syncer because the Agent may be running Client, Server, or both
simultaneously and we only want one Syncer object alive in the agent.
2016-06-10 15:54:39 -04:00
Sean Chittenden d6ef97911b
Rename Syncer.SetServiceIdentifier to SetServiceRegPrefix()
This attribute isn't actually an identifier because it can represent
a collection of services.  Rename `serviceIdentifier` to
`serviceRegPrefix which more accurately conveys the intention of this
Syncer attribute.

While here, also rename `SetServiceIdentifier()` to `SetServiceRegPrefix()`
and `GenerateServiceIdentifier()` to `GenerateServicePrefix()`.
2016-06-10 15:54:39 -04:00
Sean Chittenden 6b126ce488
Change the API signature of Syncer.SyncServices().
SyncServices() immediately attempts to sync whatever information
the process has with Consul.  Previously this method would take an
argument of the exclusive list of services that should exist,
however this is not condusive to having a Nomad Client and Nomad
Server share the same consul.Syncer.
2016-06-10 15:54:39 -04:00
Sean Chittenden fda03c5c9e
Change the signature of the PeriodicCallback to return an error
I *KNEW* I should have done this when I wrote it, but didn't want to
go back and audit the handlers to include the appropriate return
handling, but now that the code is taking shape, make this change.
2016-06-10 15:54:39 -04:00
Sean Chittenden 4973ec32bb
Rename structs.Services to structs.ConsulServices 2016-06-10 15:54:39 -04:00
Sean Chittenden e24de53ead
Refine Nomad's Consul port handling.
Previously this would immediately default to '127.0.0.1' if the
config was set to `:some-port-number`.  Now it uses the BindAddr
if available.  Also, if the `port` option is set to just a port`
number (e.g. '1234'), attempt to parse the port number by itself
to allow statically configured ports to work, even when no host is
specified.
2016-06-10 15:54:39 -04:00
Sean Chittenden 6430190d6a
Rename createCheck() to createDelegatedCheck() for clarity 2016-06-10 15:54:39 -04:00
Sean Chittenden 555f4fe135
Change client/consul.NewSyncer() to accept a shutdown channel
In addition to the API changing, consul.Syncer can now be signaled
to shutdown via the Shutdown() method, which will call the Run()'ing
sync task to exit gracefully.
2016-06-10 15:54:39 -04:00
Sean Chittenden 439baa0f8b
Remove named return parameters 2016-06-10 15:50:11 -04:00
Sean Chittenden ea0b35e303
Collapse server_endpoint_internal_test.go into server_endpoint_test.go
Requested by: @dadgar
2016-06-10 15:50:11 -04:00
Sean Chittenden 6e71bbc26b
Collapse rpcproxy_internal_test.go into rpcproxy_test.go
Requested by: @dadgar
2016-06-10 15:50:11 -04:00
Sean Chittenden 4b021b79b0
Bump the default Consul client timeout from 500ms to 5s.
Requsted by: @dadgar
2016-06-10 15:50:11 -04:00
Sean Chittenden 747a794c77
Move const block to the top of the file.
Requested by: @dadgar
2016-06-10 15:50:11 -04:00
Sean Chittenden 484816f5e0
Ensure that all accesses to Client.alloc are wrapped by allocLock. 2016-06-10 15:50:11 -04:00
Sean Chittenden 08cab4fdfa
Use client.getAllocRunners() where appropriate. 2016-06-10 15:50:11 -04:00
Sean Chittenden f9d0b9da32
Line wrap long line. 2016-06-10 15:50:11 -04:00
Sean Chittenden 0d201631a3
Rename rpcproxy.UpdateFromNodeUpdateResponse to RefreshServerLists
While breaking the API within this PR, break out the individual
arguments to RefreshServerLists.  The servers parameter is reusing
`structs.NodeServerInfo` for the time being, but this can be revisited
if the needs of the strucutre diverge in the future.
2016-06-10 15:50:11 -04:00
Sean Chittenden 0997fb1669
Fix up the comments
Pointed out by: @dadgar
2016-06-10 15:50:11 -04:00
Sean Chittenden aaa7d6bf40
Make the locking protocol more explicit in client.NewClient
With an over abundance of caution, preevnt future copy/pasta by
using the right locks when bootstrapping a Client.  Strictly speaking
this is not necessary, but it makes explicit the locking semantics
and guards against future concurrent or parallel initialization.
2016-06-10 15:50:11 -04:00
Sean Chittenden 525554c008
Use the client configCopy and lock appropriately. 2016-06-10 15:50:11 -04:00
Sean Chittenden 3060d6b33c
Flesh out the comment re: the client.rpcproxy.Run() task.
Requested by: Alex
2016-06-10 15:50:11 -04:00
Sean Chittenden a6a4345f27
Clean up various comments 2016-06-10 15:50:11 -04:00
Sean Chittenden 56a9981a13
Nuke the last of the explicit types in favor of using language idioms 2016-06-10 15:50:11 -04:00
Sean Chittenden b1ee131db8
Rename backupServerDeadline to consulPullHeartbeatDeadline
Suggested by: @alex
2016-06-10 15:50:11 -04:00
Sean Chittenden 060300007e
Use a monotonically incrementing number to create unique node names.
Also remove the space from the "name" of the node
2016-06-10 15:50:11 -04:00
Sean Chittenden 1ec7d6c266
Push down the server list even on node registration and evaluation
Be mindful of the cost of taking a snapshot from the statestore and
reuse the snapshot if one has already been taken.
2016-06-10 15:50:11 -04:00
Sean Chittenden a8d2af692c
Don't clobber the default consul config in tests 2016-06-10 15:50:11 -04:00
Sean Chittenden e2a561204a
Fix typo in comment 2016-06-10 15:50:11 -04:00
Sean Chittenden b9adfcecf5
Remove unused variable 2016-06-10 15:50:11 -04:00
Sean Chittenden f15eeb8f27
Clean up some docs and comments to be more accurate 2016-06-10 15:50:11 -04:00