4eb89481df
Adds internal endpoint read ACL support and full unit tests.
2016-05-10 11:23:47 -07:00
6a96e052c4
Adds an empty get test case.
2016-05-09 22:18:26 -07:00
471160d8f0
Performs basic plumbing of KVS transactions through all the layers.
2016-05-09 22:15:49 -07:00
dca00c96f7
Adds state store support for atomic KVS ops.
2016-05-05 15:46:59 -07:00
a1a59bee73
Splits existing KVS operations into *Txn helpers for later reuse.
2016-05-04 14:20:11 -07:00
9185450fd5
Moves KVS-related state store code out into its own set of files.
2016-05-02 16:21:04 -07:00
c16b1ca178
Add the list of Raft peers to Consul's Stats
...
```
% consul info
[snip]
raft:
[snip]
raft_peers = 127.0.0.1:8300
[snip]
```
Poached from: Nomad Project
2016-04-28 15:08:48 -07:00
79153c3014
Merge pull request #1884 from mtchavez/1541-data-dir-perms
...
command: Data directory permission error message
2016-04-12 22:06:49 -07:00
6e177a9b44
Merge pull request #1895 from shoenig/fixtypo
...
doc: fix trivial typo s/NewFSMPath/NewFSM/
2016-04-12 21:53:24 -07:00
3f340716fd
Adds a clone method to HealthCheck and uses that in local.go.
2016-04-11 00:05:39 -07:00
c9602c561c
Add description to rpc test client pool member failure message
2016-04-01 19:17:38 -07:00
7f67c123b7
doc: fix trivial typo s/NewFSMPath/NewFSM/
2016-03-29 20:52:17 -05:00
5ae7835988
Rename server_details package to agent
2016-03-29 17:39:19 -07:00
7f06c71650
Add a quick package doc for the servers package
2016-03-29 16:22:53 -07:00
897282f77d
Rename serverConfig to serverList
...
serverList is a vastly more accurate name. Chase accordingly. No functional change other than types and APIs.
2016-03-29 16:17:16 -07:00
4984b6111d
Gratuitous rename 1/2
...
Reduce cognative load and perform an overdue rename. No functional change.
Rename the `server_manager` package to `servers`. Rename the `ServerManager` package to `Manager`. In `client`, rename `serverMgr` to `servers`.
2016-03-29 16:12:00 -07:00
4734e0113f
Remove two unused constants
2016-03-29 11:11:41 -07:00
cb9833b134
Remove useless comment residual from decomposing functions
2016-03-29 10:53:00 -07:00
1f049a3c38
EDYSLEXICMOMENT
2016-03-29 10:50:10 -07:00
177f64134e
Refactor out recocileServerList anon function
...
Add testing to reconcileServerList and test various server sizes.
Test that a percentage of nodes fail their Ping (50% in testing atm)
2016-03-29 02:45:38 -07:00
6609ee5d51
Teach fauxConnPool to fail a pct of the time
...
50% failure rate seems legit as a starting point w/ 100 servers.
2016-03-28 14:53:29 -07:00
7d26f7bfa7
Call NotifyFailedServers to rotate the server list
2016-03-28 14:12:41 -07:00
6a987062b9
Add log line re: server manager backing off and sleeping
...
This is useful in situations where the RPC rotate duration is greater than 1µs. WTB exponential backoff of logging so we don't spam forever.
2016-03-28 14:04:04 -07:00
689b79aef3
Remove old debugging lines of questionable future value
2016-03-28 14:02:53 -07:00
0b0a07a280
Shuffle in place
...
Don't create a copy and save the copy, not necessary any more.
2016-03-28 14:02:27 -07:00
e230b3a3b7
Nuke unnecessary comment
...
See above function comments for details
2016-03-28 13:57:36 -07:00
34a29a2107
Move FIXME comment to the right call site
2016-03-28 13:49:55 -07:00
b38d3d71c8
Rename the ConnPoolPinger interface to Pinger
2016-03-28 13:46:01 -07:00
d6b4345375
Return error from PingConsulServer
...
In order to report why a Ping failed, change the signature of PingConsulServers to include an error message.
2016-03-28 13:38:58 -07:00
6c9fb06511
Change the definition of the ServerDetails struct key
...
Use only the serf Name for now. Leaving the plumbing for now.
2016-03-28 12:53:19 -07:00
2bcff6bac4
Correct the comment to match reality
2016-03-28 12:32:30 -07:00
fc1edea1ef
Rename serverCfg to sc for consistency
2016-03-28 12:06:26 -07:00
988b05700d
Add a quick length check
...
Verify that AddServer behaved as expected
2016-03-28 11:38:12 -07:00
7181e42ba8
Switch the order of ServerDetails.String()
...
It's more natrual to have the network first. I think I flipped the order accidentally.
2016-03-28 11:37:25 -07:00
dca8fd2643
Move rebalance log statement from INFO to DEBUG
2016-03-27 01:32:04 -07:00
180edd8e7b
Chase the API bump re: refreshServerRebalanceTimer
...
If it works in prod, why shouldn't it work in the tests?
2016-03-27 00:04:52 -07:00
9b5dd7a785
Move initialization of the rebalanceTimer to New()
2016-03-27 00:03:48 -07:00
86d1bad541
Add a test for ConnPool.PingConsulServer
...
Spin up 5x servers, join and ping each server
2016-03-26 23:52:06 -07:00
f903005080
Expose ServerManager.ResetRebalanceTimer
...
Move the rebalance timer from ServerManager.Start's stack to struct ServerManager. This makes it possible to shuffle during tests without actually waiting >120s.
2016-03-26 23:41:01 -07:00
2ba281bc5a
Logging improvements
...
Comment out noisly loggers for the time being.
Improve the final logging statement to be useful and hint what the next active server for the client is going to be.
2016-03-26 22:41:08 -07:00
fab3981b1d
Standardize the log message based on the package
...
This log statement used to belong in the consul package but has since moved to the server manager package.
2016-03-26 22:29:00 -07:00
c6d9c42d9f
Reduce the error level from Fatal when unit testing
2016-03-26 22:07:09 -07:00
4747cf3cab
Start server rebalance task after init'ing Serf
...
Now that there is no longer an event loop driven directly by Serf, start the ServerManager task after Serf has been setup. When testing and adjusting timers and timeouts to unreasonably low values, it's possible to tickle a race condition where Serf's NumNodes() would fail because Serf had not been initialized.
2016-03-26 22:04:41 -07:00
2ddf82d9d8
Catch up to a few renames
2016-03-26 19:32:11 -07:00
640ced7c11
Use empty string for addr in ServerDetails.String()
2016-03-26 19:30:04 -07:00
e0f29c17cd
Guard against a nil ServerDetails.Addr
...
It's not clear how or why this would ever be nil, but some of the unit tests produce a nil addr. Be defensive.
2016-03-26 19:29:31 -07:00
2d9982eb27
Proactively ping server before rotation
...
Before shuffling the server list, proactively ping the next server in the list to establish the connection and verify the remote endpoint is healthy.
2016-03-26 19:28:13 -07:00
b3a8e2f115
Factor out the shuffle server
2016-03-26 19:19:04 -07:00
766ddae165
Revise comments re: cycleServer
...
Improve the comments to discuss what happens presently. Add a note to consider possibly calling to TestConsulServer proactively.
2016-03-26 18:53:13 -07:00
ac1d42e9d8
Comment why the interface is needed: cyclic import
2016-03-26 18:38:35 -07:00
a9b3dba05f
Add a struct key type for server_details
2016-03-26 17:58:12 -07:00
496f05b561
Add additional checks
2016-03-25 14:40:46 -07:00
c18158aac3
Delete the right tag
...
"role" != "consul"
2016-03-25 14:31:48 -07:00
b44554f882
Don't pass in sm, server manager is already in scope
...
Go closures are implicitly capturing lambdas.
2016-03-25 14:10:09 -07:00
2713899a5b
Trim residual complexity from server join notifications
...
Now that serf node join events are decoupled from rebalancing activities completely, remove the complixity of draining the channel and ensuring only one go routine was rebalancing the server list.
Now that we're no longer initializing a notification channel, we can remove the config load/save from `Start()`
2016-03-25 14:06:35 -07:00
b3298ce4c3
Only log in FindServers
...
In FindServer this is a useful warning hinting why its call failed. RPC returns error and leaves it to the higher level caller to do whatever it wants. As an operator, I'd have the detail necessary to know why the RPC call(s) failed.
2016-03-25 13:58:50 -07:00
f024272ab2
Initialize the rebalancce to clientRPCMinReuseDuration
...
In an earlier version there was a channel to notify when a new server was added, however this has long since been removed. Just default to the sane value of 2min before the first rebalance calc takes place.
Pointed out by: slackpad
2016-03-25 13:46:18 -07:00
89311a5859
Use range vs for
...
Returning a new array vs mutating an array in place so we can use range now.
2016-03-25 13:08:08 -07:00
643997623e
Comment updates
2016-03-25 13:06:59 -07:00
072f34cf02
Only rotate server list with more than one server
...
Fantastic observation by slackpad. This was left over from when there was a boolean for health in the server struct (vs current strategy where we use server position in the list and rely on serf to cleanup the stale members).
Pointed out by: slackpad
2016-03-25 12:54:36 -07:00
aadd274a13
Relocate saveServerConfig next to getServerConfig
...
Requested by: slackpad
2016-03-25 12:41:22 -07:00
cf271e7f65
Clarify that ConsulClusterInfo is an interface over serf
...
An interface was used to break a cyclic import dependency.
2016-03-25 12:38:40 -07:00
973d924ab4
Reword comment after moving code into new packages
2016-03-25 12:34:46 -07:00
78ec9f241d
Change initialReblaanaceTimeout to a time.Duration
...
Pointed out by: @slackpad
2016-03-25 12:34:12 -07:00
328728c88a
Negative check: test an invalid condition
2016-03-25 12:22:33 -07:00
22e546ff32
Test to make sure bootstrap is missing
2016-03-25 12:20:12 -07:00
5f035da4f1
Be more Go idiomatic w/ variable names: s/valid/ok/g
...
Cargo culting is bad, m'kay?
Pointy Hat: sean-
2016-03-25 12:14:24 -07:00
e041c3905d
Fix stale comment
...
Pointed out by: @slackpad
2016-03-25 12:00:40 -07:00
45fc7c362e
Add a comment for Client serverMgr
2016-03-25 11:59:27 -07:00
5873b7e28e
Correct a bogus goimport rewrite for tests
2016-03-23 22:35:49 -07:00
dcc64d91c6
Test ServerManager.refreshServerRebalanceTimer
...
Change the signature so it returns a value so that this can be tested externally with mock data. See the sample table in TestServerManagerInternal_refreshServerRebalanceTimer() for the rate at which it will back off. This function is mostly used to not cripple large clusters in the event of a partition.
2016-03-23 22:10:50 -07:00
8e3b3d766d
Add a handful more unit tests to the public interface
2016-03-23 22:10:50 -07:00
d5f72e8c07
Rename GetNumServers to NumServers()
...
Matches the style of the rest of the repo
2016-03-23 22:10:50 -07:00
9de9cf90f1
Rename NewServerManger to just New
...
Follow go style recommendations now that this has been refactored out of the consul package and doesn't need the qualifier in the name.
2016-03-23 22:10:50 -07:00
7faea986a0
Rename FindHealthyServer() to FindServer()
...
There is no guarantee the server coming back is healthy. It's apt to be healthy by virtue of its place in the server list, but it's not guaranteed.
2016-03-23 22:10:50 -07:00
18885e3214
cycleServer is a pure function, save the result
2016-03-23 22:10:50 -07:00
4ec9ed4de2
Missed unit test cruft
2016-03-23 22:10:50 -07:00
b906e40811
Update comments to reflect reality
2016-03-23 22:10:50 -07:00
1a09a5b2cf
Remove additional cruft from ServerManager's channels
...
No longer needed code.
2016-03-23 22:10:50 -07:00
c980d492c6
Emulate a TryLock using atomic.CompareAndSwap
...
Prevent possible queueing behind serverConfigLock in the event that a server fails on a busy host.
2016-03-23 22:10:50 -07:00
102dcafe76
Make use of interfaces
...
Use an interface instead of serf.Serf as arg to NewServerManager. Bonus points for improved testability.
Pointed out by: @slackpad
2016-03-23 22:10:50 -07:00
231768faea
Simplify error handling
...
Rely on Serf for liveliness. In the event of a failure, simply cycle the server to the end of the list. If the server is unhealthy, Serf will reap the dead server.
Additional simplifications:
*) Only rebalance servers based on timers, not when a new server is readded to the cluster.
*) Back out the failure count in server_details.ServerDetails
2016-03-23 22:10:50 -07:00
0c519aa90d
Unbreak client tests by reverting to original test
...
Debugging code crept into the actual test and hung out for much longer than it should have.
2016-03-23 22:10:50 -07:00
26e51376d9
Introduce asynchronous management of consul server lists
...
Instead of blocking the RPC call path and performing a potentially expensive calculation (including a call to `c.LANMembers()`), introduce a channel to request a rebalance. Some events don't force a reshuffle, instead the extend the duration of the current rebalance window because the environment thrashed enough to redistribute a client's load.
2016-03-23 22:10:50 -07:00
6ed37d1d8d
Comment nits
2016-03-23 22:10:50 -07:00
c8ab3ae4cb
Use saveServerConfig vs atomic.Value.Store(config)
2016-03-23 22:10:50 -07:00
12377e80e6
Commit a handful of refactoring && copy/paste-o fixes
2016-03-23 22:10:50 -07:00
c1c17f158b
Mutate copies of serverCfg.servers, not original
...
Removing any ambiguity re: ownership of the mutated server lists is a win for maintenance and debugging.
2016-03-23 22:10:50 -07:00
753766cc5d
rebalanceTimer may be nil during initialization
...
When first starting the server manager, it's possible that the rebalanceTimer in serverConfig will be nil, test accordingly.
2016-03-23 22:10:50 -07:00
d0e2792d5c
Properly retain a pointer to the rebalanceTimer
2016-03-23 22:10:50 -07:00
62785de865
Cosmetic and various other wordsmithing cleanups
2016-03-23 22:10:50 -07:00
31de4290cf
Document the various functions and their locking
2016-03-23 22:10:50 -07:00
ffcd939feb
Use config convenience method to get config
...
'cause ELETTHECOMPILERSDOTHEWORK. I don't need that cluttering up the subconscious with more complexity.
2016-03-23 22:10:50 -07:00
ed7fee7a3c
Move consul.serverConfig out of the consul package
...
Relocated to its own package, server_manager. This now greatly simplifies the RPC() call path and appropriately hides the locking behind the package boundary. More work is needed to be done here
2016-03-23 22:10:50 -07:00
ab80393198
Rename serverConfigMtx to serverConfigLock
...
Pointed out by: @slackpad
2016-03-23 22:10:50 -07:00
1866d94285
Refactor out the management of Consul servers
...
Move the management of c.consulServers (fka c.consuls) into consul/server_manager.go.
This commit brings in a background task that proactively manages the server list and:
*) reshuffles the list
*) manages the timer out of the RPC() path
*) uses atomics to detect a server has failed
This is a WIP, more work in testing needs to be completed.
2016-03-23 22:10:50 -07:00
73497f7915
Move consul.serverConfig out of the consul package
...
Relocated to its own package, server_manager. This now greatly simplifies the RPC() call path and appropriately hides the locking behind the package boundary. More work is needed to be done here
2016-03-23 22:10:50 -07:00
2a52d3eb80
Rename serverConfigMtx to serverConfigLock
...
Pointed out by: @slackpad
2016-03-23 22:10:32 -07:00
49425c5371
Refactor out the management of Consul servers
...
Move the management of c.consulServers (fka c.consuls) into consul/server_manager.go.
This commit brings in a background task that proactively manages the server list and:
*) reshuffles the list
*) manages the timer out of the RPC() path
*) uses atomics to detect a server has failed
This is a WIP, more work in testing needs to be completed.
2016-03-23 22:09:46 -07:00
ebdccf0f35
Move consul.serverConfig out of the consul package
...
Relocated to its own package, server_manager. This now greatly simplifies the RPC() call path and appropriately hides the locking behind the package boundary. More work is needed to be done here
2016-03-23 22:05:29 -07:00
James Phillips