When a Nomad client that is running an allocation with
`max_client_disconnect` set misses a heartbeat the Nomad server will
update its status to `disconnected`.
Upon reconnecting, the client will make three main RPC calls:
- `Node.UpdateStatus` is used to set the client status to `ready`.
- `Node.UpdateAlloc` is used to update the client-side information about
allocations, such as their `ClientStatus`, task states etc.
- `Node.Register` is used to upsert the entire node information,
including its status.
These calls are made concurrently and are also running in parallel with
the scheduler. Depending on the order they run the scheduler may end up
with incomplete data when reconciling allocations.
For example, a client disconnects and its replacement allocation cannot
be placed anywhere else, so there's a pending eval waiting for
resources.
When this client comes back the order of events may be:
1. Client calls `Node.UpdateStatus` and is now `ready`.
2. Scheduler reconciles allocations and places the replacement alloc to
the client. The client is now assigned two allocations: the original
alloc that is still `unknown` and the replacement that is `pending`.
3. Client calls `Node.UpdateAlloc` and updates the original alloc to
`running`.
4. Scheduler notices too many allocs and stops the replacement.
This creates unnecessary placements or, in a different order of events,
may leave the job without any allocations running until the whole state
is updated and reconciled.
To avoid problems like this clients must update _all_ of its relevant
information before they can be considered `ready` and available for
scheduling.
To achieve this goal the RPC endpoints mentioned above have been
modified to enforce strict steps for nodes reconnecting:
- `Node.Register` does not set the client status anymore.
- `Node.UpdateStatus` sets the reconnecting client to the `initializing`
status until it successfully calls `Node.UpdateAlloc`.
These changes are done server-side to avoid the need of additional
coordination between clients and servers. Clients are kept oblivious of
these changes and will keep making these calls as they normally would.
The verification of whether allocations have been updates is done by
storing and comparing the Raft index of the last time the client missed
a heartbeat and the last time it updated its allocations.
Nomad server components that aren't in the `nomad` package like the deployment
watcher and volume watcher need to make RPC calls but can't import the Server
struct to do so because it creates a circular reference. These components have a
"shim" object that gets populated to pass a "static" handler that has no RPC
context.
Most RPC handlers are never used in this way, but during server setup we were
constructing a set of static handlers for most RPC endpoints anyways. This is
slightly wasteful but also confusing to developers who end up being encouraged
to just copy what was being done for previous RPCs.
This changeset includes the following refactorings:
* Remove the static handlers field on the server
* Instead construct just the specific static handlers we need to pass into the
deployment watcher and volume watcher.
* Remove the unnecessary static handler from heartbeater
* Update various tests to avoid needing the static endpoints and have them use a
endpoint constructed on the spot.
Follow-up work will examine whether we can remove the RPCs from deployment
watcher and volume watcher entirely, falling back to raft applies like node
drainer does currently.
* scheduler: allow updates after alloc reconnects
When an allocation reconnects to a cluster the scheduler needs to run
special logic to handle the reconnection, check if a replacement was
create and stop one of them.
If the allocation kept running while the node was disconnected, it will
be reconnected with `ClientStatus: running` and the node will have
`Status: ready`. This combination is the same as the normal steady state
of allocation, where everything is running as expected.
In order to differentiate between the two states (an allocation that is
reconnecting and one that is just running) the scheduler needs an extra
piece of state.
The current implementation uses the presence of a
`TaskClientReconnected` task event to detect when the allocation has
reconnected and thus must go through the reconnection process. But this
event remains even after the allocation is reconnected, causing all
future evals to consider the allocation as still reconnecting.
This commit changes the reconnect logic to use an `AllocState` to
register when the allocation was reconnected. This provides the
following benefits:
- Only a limited number of task states are kept, and they are used for
many other events. It's possible that, upon reconnecting, several
actions are triggered that could cause the `TaskClientReconnected`
event to be dropped.
- Task events are set by clients and so their timestamps are subject
to time skew from servers. This prevents using time to determine if
an allocation reconnected after a disconnect event.
- Disconnect events are already stored as `AllocState` and so storing
reconnects there as well makes it the only source of information
required.
With the new logic, the reconnection logic is only triggered if the
last `AllocState` is a disconnect event, meaning that the allocation has
not been reconnected yet. After the reconnection is handled, the new
`ClientStatus` is store in `AllocState` allowing future evals to skip
the reconnection logic.
* scheduler: prevent spurious placement on reconnect
When a client reconnects it makes two independent RPC calls:
- `Node.UpdateStatus` to heartbeat and set its status as `ready`.
- `Node.UpdateAlloc` to update the status of its allocations.
These two calls can happen in any order, and in case the allocations are
updated before a heartbeat it causes the state to be the same as a node
being disconnected: the node status will still be `disconnected` while
the allocation `ClientStatus` is set to `running`.
The current implementation did not handle this order of events properly,
and the scheduler would create an unnecessary placement since it
considered the allocation was being disconnected. This extra allocation
would then be quickly stopped by the heartbeat eval.
This commit adds a new code path to handle this order of events. If the
node is `disconnected` and the allocation `ClientStatus` is `running`
the scheduler will check if the allocation is actually reconnecting
using its `AllocState` events.
* rpc: only allow alloc updates from `ready` nodes
Clients interact with servers using three main RPC methods:
- `Node.GetAllocs` reads allocation data from the server and writes it
to the client.
- `Node.UpdateAlloc` reads allocation from from the client and writes
them to the server.
- `Node.UpdateStatus` writes the client status to the server and is
used as the heartbeat mechanism.
These three methods are called periodically by the clients and are done
so independently from each other, meaning that there can't be any
assumptions in their ordering.
This can generate scenarios that are hard to reason about and to code
for. For example, when a client misses too many heartbeats it will be
considered `down` or `disconnected` and the allocations it was running
are set to `lost` or `unknown`.
When connectivity is restored the to rest of the cluster, the natural
mental model is to think that the client will heartbeat first and then
update its allocations status into the servers.
But since there's no inherit order in these calls the reverse is just as
possible: the client updates the alloc status and then heartbeats. This
results in a state where allocs are, for example, `running` while the
client is still `disconnected`.
This commit adds a new verification to the `Node.UpdateAlloc` method to
reject updates from nodes that are not `ready`, forcing clients to
heartbeat first. Since this check is done server-side there is no need
to coordinate operations client-side: they can continue sending these
requests independently and alloc update will succeed after the heartbeat
is done.
* chagelog: add entry for #15068
* code review
* client: skip terminal allocations on reconnect
When the client reconnects with the server it synchronizes the state of
its allocations by sending data using the `Node.UpdateAlloc` RPC and
fetching data using the `Node.GetClientAllocs` RPC.
If the data fetch happens before the data write, `unknown` allocations
will still be in this state and would trigger the
`allocRunner.Reconnect` flow.
But when the server `DesiredStatus` for the allocation is `stop` the
client should not reconnect the allocation.
* apply more code review changes
* scheduler: persist changes to reconnected allocs
Reconnected allocs have a new AllocState entry that must be persisted by
the plan applier.
* rpc: read node ID from allocs in UpdateAlloc
The AllocUpdateRequest struct is used in three disjoint use cases:
1. Stripped allocs from clients Node.UpdateAlloc RPC using the Allocs,
and WriteRequest fields
2. Raft log message using the Allocs, Evals, and WriteRequest fields
3. Plan updates using the AllocsStopped, AllocsUpdated, and Job fields
Adding a new field that would only be used in one these cases (1) made
things more confusing and error prone. While in theory an
AllocUpdateRequest could send allocations from different nodes, in
practice this never actually happens since only clients call this method
with their own allocations.
* scheduler: remove logic to handle exceptional case
This condition could only be hit if, somehow, the allocation status was
set to "running" while the client was "unknown". This was addressed by
enforcing an order in "Node.UpdateStatus" and "Node.UpdateAlloc" RPC
calls, so this scenario is not expected to happen.
Adding unnecessary code to the scheduler makes it harder to read and
reason about it.
* more code review
* remove another unused test
Whenever a node joins the cluster, either for the first time or after
being `down`, we emit a evaluation for every system job to ensure all
applicable system jobs are running on the node.
This patch adds an optimization to skip creating evaluations for system
jobs not in the current node's DC. While the scheduler performs the same
feasability check, skipping the creation of the evaluation altogether
saves disk, network, and memory.
When a node fails its heart beating a number of actions are taken
to ensure state is cleaned. Service registrations a loosely tied
to nodes, therefore we should remove these from state when a node
is considered terminally down.
When a node becomes ready, create an eval for all system jobs across
namespaces.
The previous code uses `job.ID` to deduplicate evals, but that ignores
the job namespace. Thus if there are multiple jobs in different
namespaces sharing the same ID/Name, only one will be considered for
running in the new node. Thus, Nomad may skip running some system jobs
in that node.
Fix a bug where system jobs may fail to be placed on a node that
initially was not eligible for system job placement.
This changes causes the reschedule to re-evaluate the node if any
attribute used in feasibility checks changes.
Fixes https://github.com/hashicorp/nomad/issues/8448
node drain: use msgtype on txn so that events are emitted
wip: encoding extension to add Node.Drain field back to API responses
new approach for hiding Node.SecretID in the API, using `json` tag
documented this approach in the contributing guide
refactored the JSON handlers with extensions
modified event stream encoding to use the go-msgpack encoders with the extensions
This PR implements Nomad built-in support for running Consul Connect
terminating gateways. Such a gateway can be used by services running
inside the service mesh to access "legacy" services running outside
the service mesh while still making use of Consul's service identity
based networking and ACL policies.
https://www.consul.io/docs/connect/gateways/terminating-gateway
These gateways are declared as part of a task group level service
definition within the connect stanza.
service {
connect {
gateway {
proxy {
// envoy proxy configuration
}
terminating {
// terminating-gateway configuration entry
}
}
}
}
Currently Envoy is the only supported gateway implementation in
Consul. The gateay task can be customized by configuring the
connect.sidecar_task block.
When the gateway.terminating field is set, Nomad will write/update
the Configuration Entry into Consul on job submission. Because CEs
are global in scope and there may be more than one Nomad cluster
communicating with Consul, there is an assumption that any terminating
gateway defined in Nomad for a particular service will be the same
among Nomad clusters.
Gateways require Consul 1.8.0+, checked by a node constraint.
Closes#9445
* use msgtype in upsert node
adds message type to signature for upsert node, update tests, remove placeholder method
* UpsertAllocs msg type test setup
* use upsertallocs with msg type in signature
update test usage of delete node
delete placeholder msgtype method
* add msgtype to upsert evals signature, update test call sites with test setup msg type
handle snapshot upsert eval outside of FSM and ignore eval event
remove placeholder upsertevalsmsgtype
handle job plan rpc and prevent event creation for plan
msgtype cleanup upsertnodeevents
updatenodedrain msgtype
msg type 0 is a node registration event, so set the default to the ignore type
* fix named import
* fix signature ordering on upsertnode to match
properly wire up durable event count
move newline responsibility
moves newline creation from NDJson to the http handler, json stream only encodes and sends now
ignore snapshot restore if broker is disabled
enable dev mode to access event steam without acl
use mapping instead of switch
use pointers for config sizes, remove unused ttl, simplify closed conn logic
Fixes#9017
The ?resources=true query parameter includes resources in the object
stub listings. Specifically:
- For `/v1/nodes?resources=true` both the `NodeResources` and
`ReservedResources` field are included.
- For `/v1/allocations?resources=true` the `AllocatedResources` field is
included.
The ?task_states=false query parameter removes TaskStates from
/v1/allocations responses. (By default TaskStates are included.)
Add a Postrun hook to send the `CSIVolume.Unpublish` RPC to the server. This
may forward client RPCs to the node plugins or to the controller plugins,
depending on whether other allocations on this node have claims on this
volume.
By making clients responsible for running the `CSIVolume.Unpublish` RPC (and
making the RPC available to a `nomad volume detach` command), the
volumewatcher becomes only used by the core GC job and we no longer need
async volume GC from job deregister and node update.
Following the new volumewatcher in #7794 and performance improvements
to it that landed afterwards, there's no particular reason we should
be threading claim releases through the GC eval rather than writing an
empty `CSIVolumeClaimRequest` with the mode set to
`CSIVolumeClaimRelease`, just as the GC evaluation would do.
Also, by batching up these raft messages, we can reduce the amount of
raft writes by 1 and cross-server RPCs by 1 per volume we release
claims on.
Fixes#8000
When requesting a Service Identity token from Consul, use the TaskKind
of the Task to get at the service name associated with the task. In
the past using the TaskName worked because it was generated as a sidecar
task with a name that included the service. In the Native context, we
need to get at the service name in a more correct way, i.e. using the
TaskKind which is defined to include the service name.
TestClientEndpoint_CreateNodeEvals flakes a bit but its output is very
confusing, as `structs.Evaluations` overrides GoString.
Here, we emit the entire struct of the evaluation, and hopefully we'll
figure out the problem the next time it happens
This changeset adds a subsystem to run on the leader, similar to the
deployment watcher or node drainer. The `Watcher` performs a blocking
query on updates to the `CSIVolumes` table and triggers reaping of
volume claims.
This will avoid tying up scheduling workers by immediately sending
volume claim workloads into their own loop, rather than blocking the
scheduling workers in the core GC job doing things like talking to CSI
controllers
The volume watcher is enabled on leader step-up and disabled on leader
step-down.
The volume claim GC mechanism now makes an empty claim RPC for the
volume to trigger an index bump. That in turn unblocks the blocking
query in the volume watcher so it can assess which claims can be
released for a volume.
Adds a `CSIVolumeClaim` type to be tracked as current and past claims
on a volume. Allows for a client RPC failure during node or controller
detachment without having to keep the allocation around after the
first garbage collection eval.
This changeset lays groundwork for moving the actual detachment RPCs
into a volume watching loop outside the GC eval.
The `Job.Deregister` call will block on the client CSI controller RPCs
while the alloc still exists on the Nomad client node. So we need to
make the volume claim reaping async from the `Job.Deregister`. This
allows `nomad job stop` to return immediately. In order to make this
work, this changeset changes the volume GC so that the GC jobs are on a
by-volume basis rather than a by-job basis; we won't have to query
the (possibly deleted) job at the time of volume GC. We smuggle the
volume ID and whether it's a purge into the GC eval ID the same way we
smuggled the job ID previously.
* nomad/state/state_store: enforce that the volume namespace exists
* nomad/csi_endpoint_test: a couple of broken namespaces now
* nomad/csi_endpoint_test: one more test
* nomad/node_endpoint_test: use structs.DefaultNamespace
* nomad/state/state_store_test: use DefaultNamespace
* nomad/state/schema: use the namespace compound index
* scheduler/scheduler: CSIVolumeByID interface signature namespace
* scheduler/stack: SetJob on CSIVolumeChecker to capture namespace
* scheduler/feasible: pass the captured namespace to CSIVolumeByID
* nomad/state/state_store: use namespace in csi_volume index
* nomad/fsm: pass namespace to CSIVolumeDeregister & Claim
* nomad/core_sched: pass the namespace in volumeClaimReap
* nomad/node_endpoint_test: namespaces in Claim testing
* nomad/csi_endpoint: pass RequestNamespace to state.*
* nomad/csi_endpoint_test: appropriately failed test
* command/alloc_status_test: appropriately failed test
* node_endpoint_test: avoid notTheNamespace for the job
* scheduler/feasible_test: call SetJob to capture the namespace
* nomad/csi_endpoint: ACL check the req namespace, query by namespace
* nomad/state/state_store: remove deregister namespace check
* nomad/state/state_store: remove unused CSIVolumes
* scheduler/feasible: CSIVolumeChecker SetJob -> SetNamespace
* nomad/csi_endpoint: ACL check
* nomad/state/state_store_test: remove call to state.CSIVolumes
* nomad/core_sched_test: job namespace match so claim gc works
When an alloc is marked terminal (and after node unstage/unpublish
have been called), the client syncs the terminal alloc state with the
server via `Node.UpdateAlloc RPC`.
For each job that has a terminal alloc, the `Node.UpdateAlloc` RPC
handler at the server will emit an eval for a new core job to garbage
collect CSI volume claims. When this eval is handled on the core
scheduler, it will call a `volumeReap` method to release the claims
for all terminal allocs on the job.
The volume reap will issue a `ControllerUnpublishVolume` RPC for any
node that has no alloc claiming the volume. Once this returns (or
is skipped), the volume reap will send a new `CSIVolume.Claim` RPC
that releases the volume claim for that allocation in the state store,
making it available for scheduling again.
This same `volumeReap` method will be called from the core job GC,
which gives us a second chance to reclaim volumes during GC if there
were controller RPC failures.
When an alloc is marked terminal, and after node unstage/unpublish
have been called, the client will sync the terminal alloc state with
the server via `Node.UpdateAlloc` RPC.
This changeset implements releasing the volume claim for each volume
associated with the terminal alloc. It doesn't yet implement the RPC
call we need to make to the `ControllerUnpublishVolume` CSI RPC.
* state_store: csi volumes/plugins store the index in the txn
* nomad: csi_endpoint_test require index checks need uint64()
* nomad: other tests using int 0 not uint64(0)
* structs: pass index into New, but not other struct methods
* state_store: csi plugin indexes, use new struct interface
* nomad: csi_endpoint_test check index/query meta (on explicit 0)
* structs: NewCSIVolume takes an index arg now
* scheduler/test: NewCSIVolume takes an index arg now
This change updates tests to honor `BootstrapExpect` exclusively when
forming test clusters and removes test only knobs, e.g.
`config.DevDisableBootstrap`.
Background:
Test cluster creation is fragile. Test servers don't follow the
BootstapExpected route like production clusters. Instead they start as
single node clusters and then get rejoin and may risk causing brain
split or other test flakiness.
The test framework expose few knobs to control those (e.g.
`config.DevDisableBootstrap` and `config.Bootstrap`) that control
whether a server should bootstrap the cluster. These flags are
confusing and it's unclear when to use: their usage in multi-node
cluster isn't properly documented. Furthermore, they have some bad
side-effects as they don't control Raft library: If
`config.DevDisableBootstrap` is true, the test server may not
immediately attempt to bootstrap a cluster, but after an election
timeout (~50ms), Raft may force a leadership election and win it (with
only one vote) and cause a split brain.
The knobs are also confusing as Bootstrap is an overloaded term. In
BootstrapExpect, we refer to bootstrapping the cluster only after N
servers are connected. But in tests and the knobs above, it refers to
whether the server is a single node cluster and shouldn't wait for any
other server.
Changes:
This commit makes two changes:
First, it relies on `BootstrapExpected` instead of `Bootstrap` and/or
`DevMode` flags. This change is relatively trivial.
Introduce a `Bootstrapped` flag to track if the cluster is bootstrapped.
This allows us to keep `BootstrapExpected` immutable. Previously, the
flag was a config value but it gets set to 0 after cluster bootstrap
completes.
Nomad jobs may be configured with a TaskGroup which contains a Service
definition that is Consul Connect enabled. These service definitions end
up establishing a Consul Connect Proxy Task (e.g. envoy, by default). In
the case where Consul ACLs are enabled, a Service Identity token is required
for these tasks to run & connect, etc. This changeset enables the Nomad Server
to recieve RPC requests for the derivation of SI tokens on behalf of instances
of Consul Connect using Tasks. Those tokens are then relayed back to the
requesting Client, which then injects the tokens in the secrets directory of
the Task.
Copy the updated version of freeport (sdk/freeport), and tweak it for use
in Nomad tests. This means staying below port 10000 to avoid conflicts with
the lib/freeport that is still transitively used by the old version of
consul that we vendor. Also provide implementations to find ephemeral ports
of macOS and Windows environments.
Ports acquired through freeport are supposed to be returned to freeport,
which this change now also introduces. Many tests are modified to include
calls to a cleanup function for Server objects.
This should help quite a bit with some flakey tests, but not all of them.
Our port problems will not go away completely until we upgrade our vendor
version of consul. With Go modules, we'll probably do a 'replace' to swap
out other copies of freeport with the one now in 'nomad/helper/freeport'.
