This is like a Möbius strip of code due to the fact that low-level components (serf/memberlist) are connected to high-level components (the catalog and mesh-gateways) in a twisty maze of references which make it hard to dive into. With that in mind here's a high level summary of what you'll find in the patch:
There are several distinct chunks of code that are affected:
* new flags and config options for the server
* retry join WAN is slightly different
* retry join code is shared to discover primary mesh gateways from secondary datacenters
* because retry join logic runs in the *agent* and the results of that
operation for primary mesh gateways are needed in the *server* there are
some methods like `RefreshPrimaryGatewayFallbackAddresses` that must occur
at multiple layers of abstraction just to pass the data down to the right
layer.
* new cache type `FederationStateListMeshGatewaysName` for use in `proxycfg/xds` layers
* the function signature for RPC dialing picked up a new required field (the
node name of the destination)
* several new RPCs for manipulating a FederationState object:
`FederationState:{Apply,Get,List,ListMeshGateways}`
* 3 read-only internal APIs for debugging use to invoke those RPCs from curl
* raft and fsm changes to persist these FederationStates
* replication for FederationStates as they are canonically stored in the
Primary and replicated to the Secondaries.
* a special derivative of anti-entropy that runs in secondaries to snapshot
their local mesh gateway `CheckServiceNodes` and sync them into their upstream
FederationState in the primary (this works in conjunction with the
replication to distribute addresses for all mesh gateways in all DCs to all
other DCs)
* a "gateway locator" convenience object to make use of this data to choose
the addresses of gateways to use for any given RPC or gossip operation to a
remote DC. This gets data from the "retry join" logic in the agent and also
directly calls into the FSM.
* RPC (`:8300`) on the server sniffs the first byte of a new connection to
determine if it's actually doing native TLS. If so it checks the ALPN header
for protocol determination (just like how the existing system uses the
type-byte marker).
* 2 new kinds of protocols are exclusively decoded via this native TLS
mechanism: one for ferrying "packet" operations (udp-like) from the gossip
layer and one for "stream" operations (tcp-like). The packet operations
re-use sockets (using length-prefixing) to cut down on TLS re-negotiation
overhead.
* the server instances specially wrap the `memberlist.NetTransport` when running
with gateway federation enabled (in a `wanfed.Transport`). The general gist is
that if it tries to dial a node in the SAME datacenter (deduced by looking
at the suffix of the node name) there is no change. If dialing a DIFFERENT
datacenter it is wrapped up in a TLS+ALPN blob and sent through some mesh
gateways to eventually end up in a server's :8300 port.
* a new flag when launching a mesh gateway via `consul connect envoy` to
indicate that the servers are to be exposed. This sets a special service
meta when registering the gateway into the catalog.
* `proxycfg/xds` notice this metadata blob to activate additional watches for
the FederationState objects as well as the location of all of the consul
servers in that datacenter.
* `xds:` if the extra metadata is in place additional clusters are defined in a
DC to bulk sink all traffic to another DC's gateways. For the current
datacenter we listen on a wildcard name (`server.<dc>.consul`) that load
balances all servers as well as one mini-cluster per node
(`<node>.server.<dc>.consul`)
* the `consul tls cert create` command got a new flag (`-node`) to help create
an additional SAN in certs that can be used with this flavor of federation.
* xDS Mesh Gateway Resolver Subset Fixes
The first fix was that clusters were being generated for every service resolver subset regardless of there being any service instances of the associated service in that dc. The previous logic didn’t care at all but now it will omit generating those clusters unless we also have service instances that should be proxied.
The second fix was to respect the DefaultSubset of a service resolver so that mesh-gateways would configure the endpoints of the unnamed subset cluster to only those endpoints matched by the default subsets filters.
* Refactor the gateway endpoint generation to be a little easier to read
Compiling this will set an optional SNI field on each DiscoveryTarget.
When set this value should be used for TLS connections to the instances
of the target. If not set the default should be used.
Setting ExternalSNI will disable mesh gateway use for that target. It also
disables several service-resolver features that do not make sense for an
external service.
Failover is pushed entirely down to the data plane by creating envoy
clusters and putting each successive destination in a different load
assignment priority band. For example this shows that normally requests
go to 1.2.3.4:8080 but when that fails they go to 6.7.8.9:8080:
- name: foo
load_assignment:
cluster_name: foo
policy:
overprovisioning_factor: 100000
endpoints:
- priority: 0
lb_endpoints:
- endpoint:
address:
socket_address:
address: 1.2.3.4
port_value: 8080
- priority: 1
lb_endpoints:
- endpoint:
address:
socket_address:
address: 6.7.8.9
port_value: 8080
Mesh gateways route requests based solely on the SNI header tacked onto
the TLS layer. Envoy currently only lets you configure the outbound SNI
header at the cluster layer.
If you try to failover through a mesh gateway you ideally would
configure the SNI value per endpoint, but that's not possible in envoy
today.
This PR introduces a simpler way around the problem for now:
1. We identify any target of failover that will use mesh gateway mode local or
remote and then further isolate any resolver node in the compiled discovery
chain that has a failover destination set to one of those targets.
2. For each of these resolvers we will perform a small measurement of
comparative healths of the endpoints that come back from the health API for the
set of primary target and serial failover targets. We walk the list of targets
in order and if any endpoint is healthy we return that target, otherwise we
move on to the next target.
3. The CDS and EDS endpoints both perform the measurements in (2) for the
affected resolver nodes.
4. For CDS this measurement selects which TLS SNI field to use for the cluster
(note the cluster is always going to be named for the primary target)
5. For EDS this measurement selects which set of endpoints will populate the
cluster. Priority tiered failover is ignored.
One of the big downsides to this approach to failover is that the failover
detection and correction is going to be controlled by consul rather than
deferring that entirely to the data plane as with the prior version. This also
means that we are bound to only failover using official health signals and
cannot make use of data plane signals like outlier detection to affect
failover.
In this specific scenario the lack of data plane signals is ok because the
effectiveness is already muted by the fact that the ultimate destination
endpoints will have their data plane signals scrambled when they pass through
the mesh gateway wrapper anyway so we're not losing much.
Another related fix is that we now use the endpoint health from the
underlying service, not the health of the gateway (regardless of
failover mode).
In addition to exposing compilation over the API cleaned up the structures that would be exchanged to be cleaner and easier to support and understand.
Also removed ability to configure the envoy OverprovisioningFactor.
This should make them better for sending over RPC or the API.
Instead of a chain implemented explicitly like a linked list (nodes
holding pointers to other nodes) instead switch to a flat map of named
nodes with nodes linking other other nodes by name. The shipped
structure is just a map and a string to indicate which key to start
from.
Other changes:
* inline the compiler option InferDefaults as true
* introduce compiled target config to avoid needing to send back
additional maps of Resolvers; future target-specific compiled state
can go here
* move compiled MeshGateway out of the Resolver and into the
TargetConfig where it makes more sense.
* connect: reconcile how upstream configuration works with discovery chains
The following upstream config fields for connect sidecars sanely
integrate into discovery chain resolution:
- Destination Namespace/Datacenter: Compilation occurs locally but using
different default values for namespaces and datacenters. The xDS
clusters that are created are named as they normally would be.
- Mesh Gateway Mode (single upstream): If set this value overrides any
value computed for any resolver for the entire discovery chain. The xDS
clusters that are created may be named differently (see below).
- Mesh Gateway Mode (whole sidecar): If set this value overrides any
value computed for any resolver for the entire discovery chain. If this
is specifically overridden for a single upstream this value is ignored
in that case. The xDS clusters that are created may be named differently
(see below).
- Protocol (in opaque config): If set this value overrides the value
computed when evaluating the entire discovery chain. If the normal chain
would be TCP or if this override is set to TCP then the result is that
we explicitly disable L7 Routing and Splitting. The xDS clusters that
are created may be named differently (see below).
- Connect Timeout (in opaque config): If set this value overrides the
value for any resolver in the entire discovery chain. The xDS clusters
that are created may be named differently (see below).
If any of the above overrides affect the actual result of compiling the
discovery chain (i.e. "tcp" becomes "grpc" instead of being a no-op
override to "tcp") then the relevant parameters are hashed and provided
to the xDS layer as a prefix for use in naming the Clusters. This is to
ensure that if one Upstream discovery chain has no overrides and
tangentially needs a cluster named "api.default.XXX", and another
Upstream does have overrides for "api.default.XXX" that they won't
cross-pollinate against the operator's wishes.
Fixes#6159
The main change is that we no longer filter service instances by health,
preferring instead to render all results down into EDS endpoints in
envoy and merely label the endpoints as HEALTHY or UNHEALTHY.
When OnlyPassing is set to true we will force consul checks in a
'warning' state to render as UNHEALTHY in envoy.
Fixes#6171
The clusters/endpoints test were still relying on deterministic ordering of clusters/endpoints which cannot be relied upon due to golang purposefully not providing any guarantee about consistent interation ordering of maps.
Also fixed a small bug in the connect proxy cluster generation that was causing the clusters slice to be double the size it needed to with the first half being all nil pointers.
* Upgrade xDS (go-control-plane) API to support Envoy 1.10.
This includes backwards compatibility shim to work around the ext_authz package rename in 1.10.
It also adds integration test support in CI for 1.10.0.
* Fix go vet complaints
* go mod vendor
* Update Envoy version info in docs
* Update website/source/docs/connect/proxies/envoy.md
* Connect: Fix Envoy getting stuck during load
Also in this PR:
- Enabled outlier detection on upstreams which will mark instances unhealthy after 5 failures (using Envoy's defaults)
- Enable weighted load balancing where DNS weights are configured
* Fix empty load assignments in the right place
* Fix import names from review
* Move millisecond parse to a helper function
R.B. Boyer
Matt Keeler
Chris Piraino
Matt Keeler
Paul Banks