store dynamic host volume creations in client state,
so they can be "restored" on agent restart. restore works
by repeating the same Create operation as initial creation,
and expecting the plugin to be idempotent.
this is (potentially) especially important after host restarts,
which may have dropped mount points or such.
* mkdir: HostVolumePluginMkdir: just creates a directory
* example-host-volume: HostVolumePluginExternal:
plugin script that does mkfs and mount loopback
Co-authored-by: Tim Gross <tgross@hashicorp.com>
The Nomad client can now optionally emit telemetry data from the
prerun and prestart hooks. This allows operators to monitor and
alert on failures and time taken to complete.
The new datapoints are:
- nomad.client.alloc_hook.prerun.success (counter)
- nomad.client.alloc_hook.prerun.failed (counter)
- nomad.client.alloc_hook.prerun.elapsed (sample)
- nomad.client.task_hook.prestart.success (counter)
- nomad.client.task_hook.prestart.failed (counter)
- nomad.client.task_hook.prestart.elapsed (sample)
The hook execution time is useful to Nomad engineering and will
help optimize code where possible and understand job specification
impacts on hook performance.
Currently only the PreRun and PreStart hooks have telemetry
enabled, so we limit the number of new metrics being produced.
Although we have `client.allocations` metrics to track allocation states on a
client, having separate metrics for `client.tasks` will allow operators to
identify that there are individual tasks in an unexpected state in an otherwise
healthy allocation.
Fixes: https://github.com/hashicorp/nomad/issues/23770
Nomad creates Consul ACL tokens and service registrations to support Consul
service mesh workloads, before bootstrapping the Envoy proxy. Nomad always talks
to the local Consul agent and never directly to the Consul servers. But the
local Consul agent talks to the Consul servers in stale consistency mode to
reduce load on the servers. This can result in the Nomad client making the Envoy
bootstrap request with a tokens or services that have not yet replicated to the
follower that the local client is connected to. This request gets a 404 on the
ACL token and that negative entry gets cached, preventing any retries from
succeeding.
To workaround this, we'll use a method described by our friends over on
`consul-k8s` where after creating the objects in Consul we try to read them from
the local agent in stale consistency mode (which prevents a failed read from
being cached). This cannot completely eliminate this source of error because
it's possible that Consul cluster replication is unhealthy at the time we need
it, but this should make Envoy bootstrap significantly more robust.
This changset adds preflight checks for the objects we create in Consul:
* We add a preflight check for ACL tokens after we login via via Workload
Identity and in the function we use to derive tokens in the legacy
workflow. We do this check early because we also want to use this token for
registering group services in the allocrunner hooks.
* We add a preflight check for services right before we bootstrap Envoy in the
taskrunner hook, so that we have time for our service client to batch updates
to the local Consul agent in addition to the local agent sync.
We've added the timeouts to be configurable via node metadata rather than the
usual static configuration because for most cases, users should not need to
touch or even know these values are configurable; the configuration is mostly
available for testing.
Fixes: https://github.com/hashicorp/nomad/issues/9307
Fixes: https://github.com/hashicorp/nomad/issues/10451
Fixes: https://github.com/hashicorp/nomad/issues/20516
Ref: https://github.com/hashicorp/consul-k8s/pull/887
Ref: https://hashicorp.atlassian.net/browse/NET-10051
Ref: https://hashicorp.atlassian.net/browse/NET-9273
Follow-up: https://hashicorp.atlassian.net/browse/NET-10138
this is the CE side of an Enterprise-only feature.
a job trying to use this in CE will fail to validate.
to enable daily-scheduled execution entirely client-side,
a job may now contain:
task "name" {
schedule {
cron {
start = "0 12 * * * *" # may not include "," or "/"
end = "0 16" # partial cron, with only {minute} {hour}
timezone = "EST" # anything in your tzdb
}
}
...
and everything about the allocation will be placed as usual,
but if outside the specified schedule, the taskrunner will block
on the client, waiting on the schedule start, before proceeding
with the task driver execution, etc.
this includes a taksrunner hook, which watches for the end of
the schedule, at which point it will kill the task.
then, restarts-allowing, a new task will start and again block
waiting for start, and so on.
this also includes all the plumbing required to pipe API calls
through from command->api->agent->server->client, so that
tasks can be force-run, force-paused, or resume the schedule
on demand.
When `transparent_proxy` block is present and the network mode is `bridge`, use
a different CNI configuration that includes the `consul-cni` plugin. Before
invoking the CNI plugins, create a Consul SDK `iptables.Config` struct for the
allocation. This includes:
* Use all the `transparent_proxy` block fields
* The reserved ports are added to the inbound exclusion list so the alloc is
reachable from outside the mesh
* The `expose` blocks and `check` blocks with `expose=true` are added to the
inbound exclusion list so health checks work.
The `iptables.Config` is then passed as a CNI argument to the `consul-cni`
plugin.
Ref: https://github.com/hashicorp/nomad/issues/10628
* exec2: add client support for unveil filesystem isolation mode
This PR adds support for a new filesystem isolation mode, "Unveil". The
mode introduces a "alloc_mounts" directory where tasks have user-owned
directory structure which are bind mounts into the real alloc directory
structure. This enables a task driver to use landlock (and maybe the
real unveil on openbsd one day) to isolate a task to the task owned
directory structure, providing sandboxing.
* actually create alloc-mounts-dir directory
* fix doc strings about alloc mount dir paths
* exec: add a client.users configuration block
For now just add min/max dynamic user values; soon we can also absorb
the "user.denylist" and "user.checked_drivers" options from the
deprecated client.options map.
* give the no-op pool implementation a better name
* use explicit error types to make referencing them cleaner in tests
* use import alias to not shadow package name
* exec2: implement dynamic workload users taskrunner hook
This PR impelements a TR hook for allocating dynamic workload users from
a pool managed by the Nomad client. This adds a new task driver Capability,
DynamicWorkloadUsers - which a task driver must indicate in order to make
use of this feature.
The client config plumbing is coming in a followup PR - in the RFC we
realized having a client.users block would be nice to have, with some
additional unrelated options being moved from the deprecated client.options
config.
* learn to spell
This PR is the first on two that will implement the new Disconnect block. In this PR the new block is introduced to be backwards compatible with the fields it will replace. For more information refer to this RFC and this ticket.
The Nomad client expects certain cgroups paths to exist in order to
manage tasks. These paths are created when the agent first starts, but
if process fails the agent would just log the error and proceed with its
initialization, despite not being able to run tasks.
This commit surfaces the errors back to the client initialization so the
process can stop early and make clear to operators that something went
wrong.
* Move group into a separate helper module for reuse
* Add shutdownCh to worker
The shutdown channel is used to signal that worker has stopped.
* Make server shutdown block on workers' shutdownCh
* Fix waiting for eval broker state change blocking indefinitely
There was a race condition in the GenericNotifier between the
Run and WaitForChange functions, where WaitForChange blocks
trying to write to a full unsubscribeCh, but the Run function never
reads from the unsubscribeCh as it has already stopped.
This commit fixes it by unblocking if the notifier has been stopped.
* Bound the amount of time server shutdown waits on worker completion
* Fix lostcancel linter error
* Fix worker test using unexpected worker constructor
* Add changelog
---------
Co-authored-by: Marvin Chin <marvinchin@users.noreply.github.com>
If an allocrunner is persisted to the client state but the client stops before
task runner can start, we end up with an allocation in the database with
allocrunner state but no taskrunner state. This ends up mimicking an old
pre-0.9.5 state where this state was not recorded and that hits a backwards
compatibility shim. This leaves allocations in the client state that can never
be restored, but won't ever be removed either.
Update the backwards compatibility shim so that we fail the restore for the
allocrunner and remove the allocation from the client state. Taskrunners persist
state during graceful shutdown, so it shouldn't be possible to leak tasks that
have actually started. This lets us "start over" with the allocation, if the
server still wants to place it on the client.
This commit introduces the parameter preventRescheduleOnLost which indicates that the task group can't afford to have multiple instances running at the same time. In the case of a node going down, its allocations will be registered as unknown but no replacements will be rescheduled. If the lost node comes back up, the allocs will reconnect and continue to run.
In case of max_client_disconnect also being enabled, if there is a reschedule policy, an error will be returned.
Implements issue #10366
Co-authored-by: Dom Lavery <dom@circleci.com>
Co-authored-by: Tim Gross <tgross@hashicorp.com>
Co-authored-by: Luiz Aoqui <luiz@hashicorp.com>
The allocrunner has a service registration handler that proxies various API
calls to Consul. With multi-cluster support (for ENT), the service registration
handler is what selects the correct Consul client. The name of this field in the
allocrunner and taskrunner code base looks like it's referring to the actual
Consul API client. This was actually the case before Nomad native service
discovery was implemented, but now the name is misleading.
Remove the now-unused original configuration blocks for Consul and Vault from
the client. When the client needs to refer to a Consul or Vault block it will
always be for a specific cluster for the task/service. Add a helper for
accessing the default clusters (for the client's own use).
This is two of three changesets for this work. The remainder will implement the
same changes in the `command/agent` package.
As part of this work I discovered and fixed two bugs:
* The gRPC proxy socket that we create for Envoy is only ever created using the
default Consul cluster's configuration. This will prevent Connect from being
used with the non-default cluster.
* The Consul configuration we use for templates always comes from the default
Consul cluster's configuration, but will use the correct Consul token for the
non-default cluster. This will prevent templates from being used with the
non-default cluster.
Ref: https://github.com/hashicorp/nomad/issues/18947
Ref: https://github.com/hashicorp/nomad/pull/18991
Fixes: https://github.com/hashicorp/nomad/issues/18984
Fixes: https://github.com/hashicorp/nomad/issues/18983
* core: plumbing to support numa aware scheduling
* core: apply node resources compatibility upon fsm rstore
Handle the case where an upgraded server dequeus an evaluation before
a client triggers a new fingerprint - which would be needed to cause
the compatibility fix to run. By running the compat fix on restore the
server will immediately have the compatible pseudo topology to use.
* lint: learn how to spell pseudo
When agents start, they create a shared Consul client that is then wrapped as
various interfaces for testability, and used in constructing the Nomad client
and server. The interfaces that support workload services (rather than the Nomad
agent itself) need to support multiple Consul clusters for Nomad
Enterprise. Update these interfaces to be factory functions that return the
Consul client for a given cluster name. Update the `ServiceClient` to split
workload updates between clusters by creating a wrapper around all the clients
that delegates to the cluster-specific `ServiceClient`.
Ref: https://github.com/hashicorp/team-nomad/issues/404
When waiting on a previous alloc we must query against the leader before
switching to a stale query with index set.
Also check to ensure the response is fresh before using it like #18269
Similar to #18269, it is possible that even if Node.GetClientAllocs
retrieves fresh allocs that the subsequent Alloc.GetAllocs call
retrieves stale allocs. While `diffAlloc(existing, updated)` properly
ignores stale alloc *updates*, alloc deletions have no such check.
So if a client retrieves an alloc created at index 123, and then a
subsequent Alloc.GetAllocs call hits a new server which returns results
at index 100, the client will stop the alloc created at 123 because it
will be missing from the stale response.
This change applies the same logic as #18269 and ensures only fresh
responses are used.
Glossary:
* fresh - modified at an index > the query index
* stale - modified at an index <= the query index
* config: fix multi consul and vault config parse
Capture the loop variable when parsing multiple Consul and Vault
configuration blocks so the duration parse function uses the correct
field when it's called later on.
* client: build Vault client with right config
When setting up the multiple Vault clients, the code was always loading
the default configuration, resulting in all clients to be configured the
same way.
* config: fix WorkloadIdentityConfig.Copy() method
Ensure `WorkloadIdentityConfig.Copy()` does not return the original
pointer for the `TTL` field.
This commit splits identity_hook between the allocrunner and taskrunner. The
allocrunner-level part of the hook signs each task identity, and the
taskrunner-level part picks it up and stores secrets for each task.
The code revamps the WIDMgr, which is now split into 2 interfaces:
IdentityManager which manages renewals of signatures and handles sending
updates to subscribers via Watch method, and IdentitySigner which only does the
signing.
This work is necessary for having a unified Consul login workflow that comes
with the new Consul integration. A new, allocrunner-level consul_hook will now
be the only hook doing Consul authentication.
Nomad Enterprise will support configuring multiple Vault clients. Instead of
having a single Vault client field in the Nomad client, we'll have a function
that callers can parameterize by the Vault cluster name that returns the
correctly configured Vault API client wrapper.
* drivers: plumb hardware topology via grpc into drivers
This PR swaps out the temporary use of detecting system hardware manually
in each driver for using the Client's detected topology by plumbing the
data over gRPC. This ensures that Client configuration is taken to account
consistently in all references to system topology.
* cr: use enum instead of bool for core grade
* cr: fix test slit tables to be possible
* client: refactor cpuset partitioning
This PR updates the way Nomad client manages the split between tasks
that make use of resources.cpus vs. resources.cores.
Previously, each task was explicitly assigned which CPU cores they were
able to run on. Every time a task was started or destroyed, all other
tasks' cpusets would need to be updated. This was inefficient and would
crush the Linux kernel when a client would try to run ~400 or so tasks.
Now, we make use of cgroup heirarchy and cpuset inheritence to efficiently
manage cpusets.
* cr: tweaks for feedback
When restoring an allocation `WIDMgr` was not being set in the alloc
runner config, resulting in a nil panic when the task runner attempted
to start.
Since we will often require the same configuration values when creating
or restoring a new allocation, this commit moves the logic to a shared
function to ensure that `addAlloc` and `restoreState` configure alloc
runners with the same values.
Ensure that the index processed by the client is at least as new as the last index processed so that stale data does not impact the running allocations.
Allows for multiple `identity{}` blocks for tasks along with user-specified audiences. This is a building block to allow workload identities to be used with Consul, Vault and 3rd party JWT based auth methods.
Expiration is still unimplemented and is necessary for JWTs to be used securely, so that's up next.
---------
Co-authored-by: Tim Gross <tgross@hashicorp.com>
* build: update to go1.21
* go: eliminate helpers in favor of min/max
* build: run go mod tidy
* build: swap depguard for semgrep
* command: fixup broken tls error check on go1.21
In Nomad 1.5.3 we fixed a security bug that allowed bypass of ACL checks if the
request came thru a client node first. But this fix broke (knowingly) the
identification of many client-to-server RPCs. These will be now measured as if
they were anonymous. The reason for this is that many client-to-server RPCs do
not send the node secret and instead rely on the protection of mTLS.
This changeset ensures that the node secret is being sent with every
client-to-server RPC request. In a future version of Nomad we can add
enforcement on the server side, but this was left out of this changeset to
reduce risks to the safe upgrade path.
Sending the node secret as an auth token introduces a new problem during initial
introduction of a client. Clients send many RPCs concurrently with
`Node.Register`, but until the node is registered the node secret is unknown to
the server and will be rejected as invalid. This causes permission denied
errors.
To fix that, this changeset introduces a gate on having successfully made a
`Node.Register` RPC before any other RPCs can be sent (except for `Status.Ping`,
which we need earlier but which also ignores the error because that handler
doesn't do an authorization check). This ensures that we only send requests with
a node secret already known to the server. This also makes client startup a
little easier to reason about because we know `Node.Register` must succeed
first, and it should make for a good place to hook in future plans for secure
introduction of nodes. The tradeoff is that an existing client that has running
allocs will take slightly longer (a second or two) to transition to ready after
a restart, because the transition in `Node.UpdateStatus` is gated at the server
by first submitting `Node.UpdateAlloc` with client alloc updates.
This changeset adds the node pool as a label anywhere we're already emitting
labels with additional information such as node class or ID about the client.
The allocrunner sends several updates to the server during the early lifecycle
of an allocation and its tasks. Clients batch-up allocation updates every 200ms,
but experiments like the C2M challenge has shown that even with this batching,
servers can be overwhelmed with client updates during high volume
deployments. Benchmarking done in #9451 has shown that client updates can easily
represent ~70% of all Nomad Raft traffic.
Each allocation sends many updates during its lifetime, but only those that
change the `ClientStatus` field are critical for progressing a deployment or
kicking off a reschedule to recover from failures.
Add a priority to the client allocation sync and update the `syncTicker`
receiver so that we only send an update if there's a high priority update
waiting, or on every 5th tick. This means when there are no high priority
updates, the client will send updates at most every 1s instead of
200ms. Benchmarks have shown this can reduce overall Raft traffic by 10%, as
well as reduce client-to-server RPC traffic.
This changeset also switches from a channel-based collection of updates to a
shared buffer, so as to split batching from sending and prevent backpressure
onto the allocrunner when the RPC is slow. This doesn't have a major performance
benefit in the benchmarks but makes the implementation of the prioritized update
simpler.
Fixes: #9451
Tools like `nomad-nodesim` are unable to implement a minimal implementation of
an allocrunner so that we can test the client communication without having to
lug around the entire allocrunner/taskrunner code base. The allocrunner was
implemented with an interface specifically for this purpose, but there were
circular imports that made it challenging to use in practice.
Move the AllocRunner interface into an inner package and provide a factory
function type. Provide a minimal test that exercises the new function so that
consumers have some idea of what the minimum implementation required is.
When client nodes are restarted, all allocations that have been scheduled on the
node have their modify index updated, including terminal allocations. There are
several contributing factors:
* The `allocSync` method that updates the servers isn't gated on first contact
with the servers. This means that if a server updates the desired state while
the client is down, the `allocSync` races with the `Node.ClientGetAlloc`
RPC. This will typically result in the client updating the server with "running"
and then immediately thereafter "complete".
* The `allocSync` method unconditionally sends the `Node.UpdateAlloc` RPC even
if it's possible to assert that the server has definitely seen the client
state. The allocrunner may queue-up updates even if we gate sending them. So
then we end up with a race between the allocrunner updating its internal state
to overwrite the previous update and `allocSync` sending the bogus or duplicate
update.
This changeset adds tracking of server-acknowledged state to the
allocrunner. This state gets checked in the `allocSync` before adding the update
to the batch, and updated when `Node.UpdateAlloc` returns successfully. To
implement this we need to be able to equality-check the updates against the last
acknowledged state. We also need to add the last acknowledged state to the
client state DB, otherwise we'd drop unacknowledged updates across restarts.
The client restart test has been expanded to cover a variety of allocation
states, including allocs stopped before shutdown, allocs stopped by the server
while the client is down, and allocs that have been completely GC'd on the
server while the client is down. I've also bench tested scenarios where the task
workload is killed while the client is down, resulting in a failed restore.
Fixes#16381
Adds a new configuration to clients to optionally allow them to drain their
workloads on shutdown. The client sends the `Node.UpdateDrain` RPC targeting
itself and then monitors the drain state as seen by the server until the drain
is complete or the deadline expires. If it loses connection with the server, it
will monitor local client status instead to ensure allocations are stopped
before exiting.
In #16217 we switched clients using Consul discovery to the `Status.Members`
endpoint for getting the list of servers so that we're using the correct
address. This endpoint has an authorization gate, so this fails if the anonymous
policy doesn't have `node:read`. We also can't check the `AuthToken` for the
request for the client secret, because the client hasn't yet registered so the
server doesn't have anything to compare against.
Instead of hitting the `Status.Peers` or `Status.Members` RPC endpoint, use the
Consul response directly. Update the `registerNode` method to handle the list of
servers we get back in the response; if we get a "no servers" or "no path to
region" response we'll kick off discovery again and retry immediately rather
than waiting 15s.
* Update ioutil deprecated library references to os and io respectively
* Deal with the errors produced.
Add error handling to filEntry info
Add error handling to info
