Commit Graph

92 Commits

Author SHA1 Message Date
James Phillips 97e761f50f
Adds node ID integrity checking for cluster merges. 2017-03-27 00:15:42 -07:00
James Phillips 1e5a442420
Walks back the changes to change pool address interface into strings. 2017-03-16 16:42:19 -07:00
James Phillips 28f8aa5559
Removes remoteConsuls in favor of the new router.
This has the next wave of RTT integration with the router and also
factors some common RTT-related helpers out to lib. While we were
in here we also got rid of the coordinate disable config so we don't
need to deal with the complexity in the router (there was never a
user-visible way to disable coordinates).
2017-03-16 16:42:19 -07:00
James Phillips 838d85b7ae
Changes pool's dial address to a string and adds a timeout. 2017-03-16 16:42:18 -07:00
James Phillips 96bff003b7
Adds basic support for node IDs. 2017-01-17 22:47:59 -08:00
James Phillips bc29610124 Adds support for snapshots and restores. (#2396)
* Updates Raft library to get new snapshot/restore API.

* Basic backup and restore working, but need some cleanup.

* Breaks out a snapshot module and adds a SHA256 integrity check.

* Adds snapshot ACL and fills in some missing comments.

* Require a consistent read for snapshots.

* Make sure snapshot works if ACLs aren't enabled.

* Adds a bit of package documentation.

* Returns an empty response from restore to avoid EOF errors.

* Adds API client support for snapshots.

* Makes internal file names match on-disk file snapshots.

* Adds DC and token coverage for snapshot API test.

* Adds missing documentation.

* Adds a unit test for the snapshot client endpoint.

* Moves the connection pool out of the client for easier testing.

* Fixes an incidental issue in the prepared query unit test.

I realized I had two servers in bootstrap mode so this wasn't a good setup.

* Adds a half close to the TCP stream and fixes panic on error.

* Adds client and endpoint tests for snapshots.

* Moves the pool back into the snapshot RPC client.

* Adds a TLS test and fixes half-closes for TLS connections.

* Tweaks some comments.

* Adds a low-level snapshot test.

This is independent of Consul so we can pull this out into a library
later if we want to.

* Cleans up snapshot and archive and completes archive tests.

* Sends a clear error for snapshot operations in dev mode.

Snapshots require the Raft snapshots to be readable, which isn't supported
in dev mode. Send a clear error instead of a deep-down Raft one.

* Adds docs for the snapshot endpoint.

* Adds a stale mode and index feedback for snapshot saves.

This gives folks a way to extract data even if the cluster has no
leader.

* Changes the internal format of a snapshot from zip to tgz.

* Pulls in Raft fix to cancel inflight before a restore.

* Pulls in new Raft restore interface.

* Adds metadata to snapshot saves and a verify function.

* Adds basic save and restore snapshot CLI commands.

* Gets rid of tarball extensions and adds restore message.

* Fixes an incidental bad link in the KV docs.

* Adds documentation for the snapshot CLI commands.

* Scuttle any request body when a snapshot is saved.

* Fixes archive unit test error message check.

* Allows for nil output writers in snapshot RPC handlers.

* Renames hash list Decode to DecodeAndVerify.

* Closes the client connection for snapshot ops.

* Lowers timeout for restore ops.

* Updates Raft vendor to get new Restore signature and integrates with Consul.

* Bounces the leader's internal state when we do a restore.
2016-10-25 19:20:24 -07:00
Sean Chittenden 5ae7835988 Rename server_details package to agent 2016-03-29 17:39:19 -07:00
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden e041c3905d Fix stale comment
Pointed out by: @slackpad
2016-03-25 12:00:40 -07:00
Sean Chittenden 45fc7c362e Add a comment for Client serverMgr 2016-03-25 11:59:27 -07:00
Sean Chittenden d5f72e8c07 Rename GetNumServers to NumServers()
Matches the style of the rest of the repo
2016-03-23 22:10:50 -07:00
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden 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
Sean Chittenden b33648ca5c 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 16:16:22 -07:00
Sean Chittenden f3a69c939d Refactor consul.serverParts into server_details.ServerDetails
This may be short-lived, but it also seems like this is going to lead us down a path where ServerDetails is going to evolve into a more powerful package that will encapsulate more behavior behind a coherent API.
2016-03-23 16:15:47 -07:00
Sean Chittenden b3192ca410 Rename serverConfigMtx to serverConfigLock
Pointed out by: @slackpad
2016-03-23 16:15:47 -07:00
Sean Chittenden 82458fa9e8 Handle the case where there are no healthy servers
Pointed out by: @slackpad
2016-03-23 16:15:47 -07:00
Sean Chittenden 09d4c6439c 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 16:15:47 -07:00
Sean Chittenden 7de85906c1 Rename `lastServer` to `preferredServer`
Expanding the domain of lastServer beyond RPC() changes the meaning of this variable.  Rename accordingly to match the intent coming in a subsequent commit: a background thread will be in charge of rotating preferredServer.
2016-03-23 16:14:59 -07:00
Sean Chittenden 2949980a64 Warn if serf events have queued up past 80% of the limit
It is theoretically possible that the number of queued serf events can back up.  If this happens, emit a warning message if there are more than 200 events in queue.

Most notably, this can happen if `c.consulServerLock` is held for an "extended period of time".  The probability of anyone ever seeing this log message is hopefully low to nonexistent, but if it happens, the warning message indicating a large number of serf events fired while a lock was held is likely to be helpful (vs serf mysteriously blocking when attempting to add an event to a channel).
2016-03-23 16:14:11 -07:00
Sean Chittenden 3ac1bcc799 Remove lastRPCTime
This mechanism isn't going to provide much value in the future.  Preemptively reduce the complexity of future work.
2016-03-23 16:13:49 -07:00
Sean Chittenden 72b7856045 Rename c.consuls to c.consulServers
Prep for breaking out maintenance of consuls into a new goroutine.
2016-03-23 16:10:27 -07:00
James Phillips d660311fbb Revert "Merge pull request #1667 from hashicorp/b-redistribute-clients"
This reverts commit 8f30dea4209491ebbe4ef9ab94dd8052d17bdbe9, reversing
changes made to eb27a02956e7e052c0bec6f96a0c0f7f6675f6a6.
2016-02-24 15:38:03 -08:00
Sean Chittenden fc82b351b8 Use the server's address in debug logging, not the c.lastServer, which may be nil 2016-02-02 15:51:28 -08:00
Sean Chittenden 58225e0ee3 Remove unnecessary check, test was moved further up in scope 2016-02-02 11:13:58 -08:00
Sean Chittenden ef8bbca48f Continually rebalance client connections
Introduce a low-level background connection expiration mechanism wherein connections will be recycled periodically based on the size and health of the cluster.

For the vast majority of consul users, this will mean an average connection age of 150s.  For 10K node clusters it will take ~3min for clusters to rebalance their connections.  In the pathological case for a 100K cluster where 99K clients are in the minority talking to 1x server it will take ~26min to rebalance all connections.

It's possibe for clients recovering from a parititon to become fixated on a single server until the server or agent is restarted.  This is of particular interest to long-running environments with stable agents, where `allow_stale` is true, and partitions occur periodically.
2016-01-30 17:13:50 -08:00
Sean Chittenden 8a37e76cb0 Use rand.Int31n() vs unconditionally using modulus 2016-01-30 15:47:58 -08:00
Sean Chittenden b216d4c11f Rename clientRPCCache to clientRPCConnMaxIdle, change value
Increase the max idle time for agents talking to servers from 30s to 127s in order to allow for the reuse of connections that are being initiated by cron.

127s was chosen as the first prime above 120s (arbitrarily chose to use a prime) with the intent of reusing connections who are used by once-a-minute cron(8) jobs *and* who use a 60s jitter window (e.g. in vixie cron job execution can drift by up to 59s per job, or 119s for a once-a-minute cron job).
2016-01-30 15:27:46 -08:00
Sean Chittenden e83a5b7a70 Reuse the results from gettimeofday(2)...
Inside of a single RPC call, reuse time.Now().
2016-01-30 14:39:17 -08:00
James Phillips 132e1d813b Fixes configs now that Serf always caches coordinates. 2015-10-23 15:23:01 -07:00
James Phillips f71c79c53f Does some small cleanups based on PR feedback.
* Holds coordinate updates in map and gets rid of the update channel.
* Cleans up config variables a bit.
2015-10-23 15:23:01 -07:00
James Phillips b6c31bdf2f Flips the sense of the coordinate enable option. 2015-10-23 15:23:01 -07:00
James Phillips edb9a119e2 Does a clean up pass on the Consul side. 2015-10-23 15:23:01 -07:00
Derek Chiang eb599a1745 Address comments 2015-10-23 15:23:01 -07:00
Derek Chiang b2cff43bb5 Complete logic for sending coordinates 2015-10-23 15:23:01 -07:00
Dale Wijnand c5168e1263 Fix a bunch of typos. 2015-09-15 13:22:08 +01:00
Ryan Uber e6923a4832 consul: always fire events from server nodes 2015-06-18 18:13:29 -07:00
Sam Boyer bdc5983463 Condense switch fallthroughs into expr lists 2015-05-26 21:30:14 -04:00
Armon Dadgar 9642384429 consul: support the new TLS wrapper 2015-05-11 15:15:36 -07:00
Armon Dadgar 3bf337a6ac consul: thread the target DC through the RPC path 2015-05-11 13:09:19 -07:00
Armon Dadgar a1de4b17c2 consul: use tlsutil.Wrapper instead of tls.Config directly 2015-05-11 13:09:19 -07:00
Ryan Uber 275d99e1dc consul: allow returning custom error for merge delegate 2015-02-22 18:24:10 -08:00
Armon Dadgar a66a765ca9 consul: Adding merge delegate to prevent mixing clusters 2015-01-06 15:48:46 -08:00
Ryan Uber 4cd89a9113 Rebase against upstream 2014-11-19 16:45:49 -08:00
Ryan Uber 2661bbfa27 consul: more tests, remove unused KeyManager() method 2014-11-19 16:37:40 -08:00