…rvices (#10893)

The "all mirror services have endpoints" check can fail in the presence
of lots of mirrored services because for each service we query the kube
api for its endpoints, and those calls reuse the same golang context,
which ends up reaching its deadline.

To fix, we create a new context object per call.

## Repro

First patch `check.go` to introduce a sleep in order to simulate network
latency:

```diff
diff --git a/multicluster/cmd/check.go b/multicluster/cmd/check.go
index b2b4158bf..f3083f436 100644
--- a/multicluster/cmd/check.go
+++ b/multicluster/cmd/check.go
@@ -627,6 +627,7 @@ func (hc *healthChecker) checkIfMirrorServicesHaveEndpoints(ctx context.Context)
        for _, svc := range mirrorServices.Items {
                // Check if there is a relevant end-point
                endpoint, err := hc.KubeAPIClient().CoreV1().Endpoints(svc.Namespace).Get(ctx, svc.Name, metav1.GetOptions{})
+               time.Sleep(1 * time.Second)
                if err != nil || len(endpoint.Subsets) == 0 {
                        servicesWithNoEndpoints = append(servicesWithNoEndpoints, fmt.Sprintf("%s.%s mirrored from cluster [%s]", svc.Name, svc.Namespace, svc.Labels[k8s.RemoteClusterNameLabel]))
                }
```

Then run the `multicluster` integration tests to setup a multicluster
scenario, and then create lots of mirrored services!

```bash
$ bin/docker-build

# accommodate to your own arch
$ bin/tests --name multicluster --skip-cluster-delete $PWD/target/cli/linux-amd64/linkerd

# we are currently in the target cluster context
$ k create ns testing

# create pod
$ k -n testing run nginx --image=nginx --restart=Never

# create 50 services pointing to it, flagged to be mirrored
$ for i in {1..50}; do k -n testing expose po nginx --port 80 --name "nginx-$i" -l mirror.linkerd.io/exported=true; done

# switch to the source cluster
$ k config use-context k3d-source

# this will trigger the creation of the mirrored services, wait till the
# 50 are created
$ k create ns testing

$ bin/go-run cli mc check --verbose
github.com/linkerd/linkerd2/multicluster/cmd
github.com/linkerd/linkerd2/cli/cmd
linkerd-multicluster
--------------------
√ Link CRD exists
√ Link resources are valid
        * target
√ remote cluster access credentials are valid
        * target
√ clusters share trust anchors
        * target
√ service mirror controller has required permissions
        * target
√ service mirror controllers are running
        * target
DEBU[0000] Starting port forward to https://0.0.0.0:34201/api/v1/namespaces/linkerd-multicluster/pods/linkerd-service-mirror-target-7c4496869f-6xsp4/portforward?timeout=30s 39327:9999
DEBU[0000] Port forward initialised
√ probe services able to communicate with all gateway mirrors
        * target
DEBU[0031] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0032] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0033] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0034] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0035] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0036] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
DEBU[0037] error retrieving Endpoints: client rate limiter Wait returned an error: context deadline exceeded
```

It's useful to modify values for `CheckURL` and `DefaultDockerRegistry`,
to test different configurations of Linkerd. To modify these at build
time, they must be public vars.

Make `CheckURL` and `DefaultDockerRegistry` public vars, to allow for
build-time modification using Go's `-ldflags`.

Also, DRY up `DefaultDockerRegistry`.

Signed-off-by: Andrew Seigner <[email protected]>

The `linkerd check --output` flag supports 3 formats: `table`, `json`,
and `short`. The default `linkerd check` command description incorrectly
printed `basic, json, short`. Other extension check commands printed
`basic, json`.

Modify all check output descriptions to print `table, json, short`.

Signed-off-by: Andrew Seigner <[email protected]>

The `--log-level` flag did not support a `trace` level, despite the
underlying `logrus` library supporting it. Also, at `debug` level, the
Control Plane components were setting klog at v=12, which includes
sensitive data.

Introduce a `trace` log level. Keep klog at `v=12` for `trace`, change
it to `v=6` for `debug`.

Fixes #11132

Signed-off-by: Andrew Seigner <[email protected]>

Fixes #11163

The `servicePublisher.updateServer` function will iterate through all registered listeners and update them.  However, a nil listener may temporarily be in the list of listeners if an unsubscribe is in progress.  This results in a nil pointer dereference.

All functions which result in updating the listeners must therefore be protected by the mutex so that we don't try to act on the list of listeners while it is being modified.

Signed-off-by: Alex Leong <[email protected]>

Problem:
Commands `jaeger install`, `multicluster link` give precedence to `LINKERD_DOCKER_REGISTRY` env var, whereas commands `install`, `upgrade` and `inject` give preference to `--register` flag.

Solution:
Make the commands consitent by giving precedence to `--register` flag in all commands.

Fixes: #11115

Signed-off-by: Harsh Soni <[email protected]>

Problem - Current does Linkerd CNI Helm chart templates have hostNetwork: true set which is unnecessary and less secure.

Solution - Removed hostNetwork: true from linkerd-cni Helm chart templates

PR Fixes #11141 
---------

Signed-off-by: Abhijeet Gaurav <[email protected]>
Co-authored-by: Alejandro Pedraza <[email protected]>

In 2.13, the default inbound and outbound HTTP request queue capacity
decreased from 10,000 requests to 100 requests (in PR #2078). This
change results in proxies shedding load much more aggressively while
under high load to a single destination service, resulting in increased
error rates in comparison to 2.12 (see #11055 for
details).

This commit changes the default HTTP request queue capacities for the
inbound and outbound proxies back to 10,000 requests, the way they were
in 2.12 and earlier. In manual load testing I've verified that
increasing the queue capacity results in a substantial decrease in 503
Service Unavailable errors emitted by the proxy: with a queue capacity
of 100 requests, the load test described [here] observed a failure rate
of 51.51% of requests, while with a queue capacity of 10,000 requests,
the same load test observes no failures.

Note that I did not modify the TCP connection queue capacities, or the
control plane request queue capacity. These were previously configured
by the same variable before #2078, but were split out into separate vars
in that change. I don't think the queue capacity limits for TCP
connection establishment or for control plane requests are currently
resulting in instability the way the decreased request queue capacity
is, so I decided to make a more focused change to just the HTTP request
queues for the proxies.

[here]: #11055 (comment)

---

* Increase HTTP request queue capacity (linkerd/linkerd2-proxy#2449)

This release stops using the "interface" mode, and instead wait till
another CNI plugin drops a proper network config and then append the
linkerd CNI config to it. This avoids having pods start before proper
networking is established in the node.

This stable release fixes a regression introduced in stable-2.13.0 which
resulted in proxies shedding load too aggressively while under moderate
request load to a single service ([#11055]). In addition, it updates the
base image for the `linkerd-cni` initcontainer to resolve a CVE in
`libdb` ([#11196]), fixes a race condition in the Destination controller
that could cause it to crash ([#11163]), as well as fixing a number of
other issues.

* Control Plane
  * Fixed a race condition in the destination controller that could
    cause it to panic ([#11169]; fixes [#11193])
  * Improved the granularity of logging levels in the control plane
    ([#11147])

* Proxy
  * Changed the default HTTP request queue capacities for the inbound
    and outbound proxies back to 10,000 requests ([#11198]; fixes
    [#11055])

* CLI
  * Updated extension CLI commands to prefer the `--registry` flag over
    the `LINKERD_DOCKER_REGISTRY` environment variable, making the
    precedence more consistent (thanks @harsh020!) (see [#11144])

* CNI
  * Updated `linkerd-cni` base image to resolve [CVE-2019-8457] in
    `libdb` ([#11196])
  * Changed the CNI plugin installer to always run in 'chained' mode;
    the plugin will now wait until another CNI plugin is installed
    before appending its configuration ([#10849])
  * Removed `hostNetwork: true` from linkerd-cni Helm chart templates
    ([#11158]; fixes [#11141]) (thanks @abhijeetgauravm!)

* Multicluster
  * Fixed the `linkerd multicluster check` command failing in the
    presence of lots of mirrored services ([#10764])

[#10764]: #10764
[#10849]: #10849
[#11055]: #11055
[#11141]: #11141
[#11144]: #11144
[#11147]: #11147
[#11158]: #11158
[#11163]: #11163
[#11169]: #11169
[#11196]: #11196
[#11198]: #11198
[CVE-2019-8457]: https://avd.aquasec.com/nvd/2019/cve-2019-8457/