consul/test/integration/connect/envoy/helpers.bash

940 lines
24 KiB
Bash
Raw Normal View History

#!/bin/bash
# retry based on
# https://github.com/fernandoacorreia/azure-docker-registry/blob/master/tools/scripts/create-registry-server
# under MIT license.
function retry {
local n=1
local max=$1
shift
local delay=$1
shift
local errtrace=0
if grep -q "errtrace" <<< "$SHELLOPTS"
then
errtrace=1
set +E
fi
for ((i=1;i<=$max;i++))
do
if "$@"
then
if test $errtrace -eq 1
then
set -E
fi
return 0
else
echo "Command failed. Attempt $i/$max:"
sleep $delay
fi
done
if test $errtrace -eq 1
then
set -E
fi
return 1
}
function retry_default {
set +E
ret=0
retry 5 1 "$@" || ret=1
set -E
return $ret
}
function retry_long {
retry 30 1 "$@"
}
function echored {
tput setaf 1
tput bold
echo $@
tput sgr0
}
function echogreen {
tput setaf 2
tput bold
echo $@
tput sgr0
}
function echoyellow {
tput setaf 3
tput bold
echo $@
tput sgr0
}
function echoblue {
tput setaf 4
tput bold
echo $@
tput sgr0
}
function is_set {
# Arguments:
# $1 - string value to check its truthiness
#
# Return:
# 0 - is truthy (backwards I know but allows syntax like `if is_set <var>` to work)
# 1 - is not truthy
local val=$(tr '[:upper:]' '[:lower:]' <<< "$1")
case $val in
1 | t | true | y | yes)
return 0
;;
*)
return 1
;;
esac
}
function get_cert {
local HOSTPORT=$1
local SERVER_NAME=$2
local CA_FILE=$3
local SNI_FLAG=""
if [ -n "$SERVER_NAME" ]; then
SNI_FLAG="-servername $SERVER_NAME"
fi
CERT=$(openssl s_client -connect $HOSTPORT $SNI_FLAG -showcerts </dev/null)
openssl x509 -noout -text <<< "$CERT"
}
function assert_proxy_presents_cert_uri {
local HOSTPORT=$1
local SERVICENAME=$2
local DC=${3:-primary}
local NS=${4:-default}
2021-11-12 21:45:50 +00:00
local PARTITION=${5:default}
CERT=$(retry_default get_cert $HOSTPORT)
2021-11-12 21:45:50 +00:00
echo "WANT SERVICE: ${PARTITION}/${NS}/${SERVICENAME}"
echo "GOT CERT:"
echo "$CERT"
2021-11-12 21:45:50 +00:00
if [[ -z $PARTITION ]] || [[ $PARTITION = "default" ]]; then
echo "$CERT" | grep -Eo "URI:spiffe://([a-zA-Z0-9-]+).consul/ns/${NS}/dc/${DC}/svc/$SERVICENAME"
else
echo "$CERT" | grep -Eo "URI:spiffe://([a-zA-Z0-9-]+).consul/ap/${PARTITION}/ns/${NS}/dc/${DC}/svc/$SERVICENAME"
fi
}
function assert_dnssan_in_cert {
local HOSTPORT=$1
local DNSSAN=$2
local SERVER_NAME=${3:-$DNSSAN}
CERT=$(retry_default get_cert $HOSTPORT $SERVER_NAME)
echo "WANT DNSSAN: ${DNSSAN} (SNI: ${SERVER_NAME})"
echo "GOT CERT:"
echo "$CERT"
echo "$CERT" | grep -Eo "DNS:${DNSSAN}"
}
function assert_cert_signed_by_ca {
local CA_FILE=$1
local HOSTPORT=$2
local DNSSAN=$3
local SERVER_NAME=${4:-$DNSSAN}
local SNI_FLAG=""
if [ -n "$SERVER_NAME" ]; then
SNI_FLAG="-servername $SERVER_NAME"
fi
CERT=$(openssl s_client -connect $HOSTPORT $SNI_FLAG -CAfile $CA_FILE -showcerts </dev/null)
echo "GOT CERT:"
echo "$CERT"
echo "$CERT" | grep 'Verify return code: 0 (ok)'
}
function assert_envoy_version {
local ADMINPORT=$1
run retry_default curl -f -s localhost:$ADMINPORT/server_info
[ "$status" -eq 0 ]
# Envoy 1.8.0 returns a plain text line like
# envoy 5d25f466c3410c0dfa735d7d4358beb76b2da507/1.8.0/Clean/DEBUG live 3 3 0
# Later versions return JSON.
if (echo $output | grep '^envoy') ; then
VERSION=$(echo $output | cut -d ' ' -f 2)
else
VERSION=$(echo $output | jq -r '.version')
fi
echo "Status=$status"
echo "Output=$output"
echo "---"
echo "Got version=$VERSION"
echo "Want version=$ENVOY_VERSION"
# 1.20.2 is a special snowflake in that the version for the release is actually
# reported as '1.20.2-dev'
if [ "$ENVOY_VERSION" = "1.20.2" ] ; then
ENVOY_VERSION="1.20.2-dev"
fi
echo $VERSION | grep "/$ENVOY_VERSION/"
}
function assert_envoy_expose_checks_listener_count {
local HOSTPORT=$1
local EXPECT_PATH=$2
# scrape this once
BODY=$(get_envoy_expose_checks_listener_once $HOSTPORT)
echo "BODY = $BODY"
CHAINS=$(echo "$BODY" | jq '.active_state.listener.filter_chains | length')
echo "CHAINS = $CHAINS (expect 1)"
[ "${CHAINS:-0}" -eq 1 ]
RANGES=$(echo "$BODY" | jq '.active_state.listener.filter_chains[0].filter_chain_match.source_prefix_ranges | length')
echo "RANGES = $RANGES (expect 3)"
# note: if IPv6 is not supported in the kernel per
# agent/xds:kernelSupportsIPv6() then this will only be 2
[ "${RANGES:-0}" -eq 3 ]
HCM=$(echo "$BODY" | jq '.active_state.listener.filter_chains[0].filters[0]')
HCM_NAME=$(echo "$HCM" | jq -r '.name')
HCM_PATH=$(echo "$HCM" | jq -r '.typed_config.route_config.virtual_hosts[0].routes[0].match.path')
echo "HCM = $HCM"
[ "${HCM_NAME:-}" == "envoy.filters.network.http_connection_manager" ]
[ "${HCM_PATH:-}" == "${EXPECT_PATH}" ]
}
function get_envoy_expose_checks_listener_once {
local HOSTPORT=$1
run curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '.configs[] | select(.["@type"] == "type.googleapis.com/envoy.admin.v3.ListenersConfigDump") | .dynamic_listeners[] | select(.name | startswith("exposed_path_"))'
}
function assert_envoy_http_rbac_policy_count {
local HOSTPORT=$1
local EXPECT_COUNT=$2
GOT_COUNT=$(get_envoy_http_rbac_once $HOSTPORT | jq '.rules.policies | length')
echo "GOT_COUNT = $GOT_COUNT"
[ "${GOT_COUNT:-0}" -eq $EXPECT_COUNT ]
}
function get_envoy_http_rbac_once {
local HOSTPORT=$1
run curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '.configs[2].dynamic_listeners[].active_state.listener.filter_chains[0].filters[0].typed_config.http_filters[] | select(.name == "envoy.filters.http.rbac") | .typed_config'
}
function assert_envoy_network_rbac_policy_count {
local HOSTPORT=$1
local EXPECT_COUNT=$2
GOT_COUNT=$(get_envoy_network_rbac_once $HOSTPORT | jq '.rules.policies | length')
echo "GOT_COUNT = $GOT_COUNT"
[ "${GOT_COUNT:-0}" -eq $EXPECT_COUNT ]
}
function get_envoy_network_rbac_once {
local HOSTPORT=$1
run curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '.configs[2].dynamic_listeners[].active_state.listener.filter_chains[0].filters[] | select(.name == "envoy.filters.network.rbac") | .typed_config'
}
function get_envoy_listener_filters {
local HOSTPORT=$1
run retry_default curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '.configs[2].dynamic_listeners[].active_state.listener | "\(.name) \( .filter_chains[0].filters | map(.name) | join(","))"'
}
function get_envoy_http_filters {
local HOSTPORT=$1
run retry_default curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '.configs[2].dynamic_listeners[].active_state.listener | "\(.name) \( .filter_chains[0].filters[] | select(.name == "envoy.filters.network.http_connection_manager") | .typed_config.http_filters | map(.name) | join(","))"'
}
Support Incremental xDS mode (#9855) This adds support for the Incremental xDS protocol when using xDS v3. This is best reviewed commit-by-commit and will not be squashed when merged. Union of all commit messages follows to give an overarching summary: xds: exclusively support incremental xDS when using xDS v3 Attempts to use SoTW via v3 will fail, much like attempts to use incremental via v2 will fail. Work around a strange older envoy behavior involving empty CDS responses over incremental xDS. xds: various cleanups and refactors that don't strictly concern the addition of incremental xDS support Dissolve the connectionInfo struct in favor of per-connection ResourceGenerators instead. Do a better job of ensuring the xds code uses a well configured logger that accurately describes the connected client. xds: pull out checkStreamACLs method in advance of a later commit xds: rewrite SoTW xDS protocol tests to use protobufs rather than hand-rolled json strings In the test we very lightly reuse some of the more boring protobuf construction helper code that is also technically under test. The important thing of the protocol tests is testing the protocol. The actual inputs and outputs are largely already handled by the xds golden output tests now so these protocol tests don't have to do double-duty. This also updates the SoTW protocol test to exclusively use xDS v2 which is the only variant of SoTW that will be supported in Consul 1.10. xds: default xds.Server.AuthCheckFrequency at use-time instead of construction-time
2021-04-29 18:54:05 +00:00
function get_envoy_dynamic_cluster_once {
local HOSTPORT=$1
local NAME_PREFIX=$2
run curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output ".configs[] | select (.[\"@type\"] == \"type.googleapis.com/envoy.admin.v3.ClustersConfigDump\") | .dynamic_active_clusters[] | select(.cluster.name | startswith(\"${NAME_PREFIX}\"))"
}
function assert_envoy_dynamic_cluster_exists_once {
local HOSTPORT=$1
local NAME_PREFIX=$2
local EXPECT_SNI=$3
BODY="$(get_envoy_dynamic_cluster_once $HOSTPORT $NAME_PREFIX)"
[ -n "$BODY" ]
SNI="$(echo "$BODY" | jq --raw-output ".cluster.transport_socket.typed_config.sni | select(. | startswith(\"${EXPECT_SNI}\"))")"
[ -n "$SNI" ]
}
function assert_envoy_dynamic_cluster_exists {
local HOSTPORT=$1
local NAME_PREFIX=$2
local EXPECT_SNI=$3
run retry_long assert_envoy_dynamic_cluster_exists_once $HOSTPORT $NAME_PREFIX $EXPECT_SNI
[ "$status" -eq 0 ]
}
function get_envoy_cluster_config {
local HOSTPORT=$1
local CLUSTER_NAME=$2
run retry_default curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output "
.configs[1].dynamic_active_clusters[]
| select(.cluster.name|startswith(\"${CLUSTER_NAME}\"))
| .cluster
"
}
function get_envoy_stats_flush_interval {
local HOSTPORT=$1
run retry_default curl -s -f $HOSTPORT/config_dump
[ "$status" -eq 0 ]
#echo "$output" > /workdir/s1_envoy_dump.json
echo "$output" | jq --raw-output '.configs[0].bootstrap.stats_flush_interval'
}
# snapshot_envoy_admin is meant to be used from a teardown scriptlet from the host.
function snapshot_envoy_admin {
local HOSTPORT=$1
local ENVOY_NAME=$2
local DC=${3:-primary}
local OUTDIR="${LOG_DIR}/envoy-snapshots/${DC}/${ENVOY_NAME}"
mkdir -p "${OUTDIR}"
docker_wget "$DC" "http://${HOSTPORT}/config_dump" -q -O - > "${OUTDIR}/config_dump.json"
docker_wget "$DC" "http://${HOSTPORT}/clusters?format=json" -q -O - > "${OUTDIR}/clusters.json"
docker_wget "$DC" "http://${HOSTPORT}/stats" -q -O - > "${OUTDIR}/stats.txt"
docker_wget "$DC" "http://${HOSTPORT}/stats/prometheus" -q -O - > "${OUTDIR}/stats_prometheus.txt"
}
function reset_envoy_metrics {
local HOSTPORT=$1
curl -s -f -XPOST $HOSTPORT/reset_counters
return $?
}
function get_all_envoy_metrics {
local HOSTPORT=$1
curl -s -f $HOSTPORT/stats
return $?
}
function get_envoy_metrics {
local HOSTPORT=$1
local METRICS=$2
get_all_envoy_metrics $HOSTPORT | grep "$METRICS"
}
function get_upstream_endpoint_in_status_count {
local HOSTPORT=$1
local CLUSTER_NAME=$2
local HEALTH_STATUS=$3
run curl -s -f "http://${HOSTPORT}/clusters?format=json"
[ "$status" -eq 0 ]
# echo "$output" >&3
echo "$output" | jq --raw-output "
.cluster_statuses[]
| select(.name|startswith(\"${CLUSTER_NAME}\"))
| [.host_statuses[].health_status.eds_health_status]
| [select(.[] == \"${HEALTH_STATUS}\")]
| length"
}
function assert_upstream_has_endpoints_in_status_once {
local HOSTPORT=$1
local CLUSTER_NAME=$2
local HEALTH_STATUS=$3
local EXPECT_COUNT=$4
GOT_COUNT=$(get_upstream_endpoint_in_status_count $HOSTPORT $CLUSTER_NAME $HEALTH_STATUS)
[ "$GOT_COUNT" -eq $EXPECT_COUNT ]
}
function assert_upstream_has_endpoints_in_status {
local HOSTPORT=$1
local CLUSTER_NAME=$2
local HEALTH_STATUS=$3
local EXPECT_COUNT=$4
run retry_long assert_upstream_has_endpoints_in_status_once $HOSTPORT $CLUSTER_NAME $HEALTH_STATUS $EXPECT_COUNT
[ "$status" -eq 0 ]
}
function assert_envoy_metric {
set -eEuo pipefail
local HOSTPORT=$1
local METRIC=$2
local EXPECT_COUNT=$3
METRICS=$(get_envoy_metrics $HOSTPORT "$METRIC")
if [ -z "${METRICS}" ]
then
echo "Metric not found" 1>&2
return 1
fi
GOT_COUNT=$(awk -F: '{print $2}' <<< "$METRICS" | head -n 1 | tr -d ' ')
if [ -z "$GOT_COUNT" ]
then
echo "Couldn't parse metric count" 1>&2
return 1
fi
if [ $EXPECT_COUNT -ne $GOT_COUNT ]
then
echo "$METRIC - expected count: $EXPECT_COUNT, actual count: $GOT_COUNT" 1>&2
return 1
fi
}
function assert_envoy_metric_at_least {
set -eEuo pipefail
local HOSTPORT=$1
local METRIC=$2
local EXPECT_COUNT=$3
METRICS=$(get_envoy_metrics $HOSTPORT "$METRIC")
if [ -z "${METRICS}" ]
then
echo "Metric not found" 1>&2
return 1
fi
GOT_COUNT=$(awk -F: '{print $2}' <<< "$METRICS" | head -n 1 | tr -d ' ')
if [ -z "$GOT_COUNT" ]
then
echo "Couldn't parse metric count" 1>&2
return 1
fi
if [ $EXPECT_COUNT -gt $GOT_COUNT ]
then
echo "$METRIC - expected >= count: $EXPECT_COUNT, actual count: $GOT_COUNT" 1>&2
return 1
fi
}
function assert_envoy_aggregate_metric_at_least {
set -eEuo pipefail
local HOSTPORT=$1
local METRIC=$2
local EXPECT_COUNT=$3
METRICS=$(get_envoy_metrics $HOSTPORT "$METRIC")
if [ -z "${METRICS}" ]
then
echo "Metric not found" 1>&2
return 1
fi
GOT_COUNT=$(awk '{ sum += $2 } END { print sum }' <<< "$METRICS")
if [ -z "$GOT_COUNT" ]
then
echo "Couldn't parse metric count" 1>&2
return 1
fi
if [ $EXPECT_COUNT -gt $GOT_COUNT ]
then
echo "$METRIC - expected >= count: $EXPECT_COUNT, actual count: $GOT_COUNT" 1>&2
return 1
fi
}
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
function get_healthy_service_count {
local SERVICE_NAME=$1
connect: fix failover through a mesh gateway to a remote datacenter (#6259) 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).
2019-08-05 18:30:35 +00:00
local DC=$2
local NS=$3
run curl -s -f ${HEADERS} "127.0.0.1:8500/v1/health/connect/${SERVICE_NAME}?dc=${DC}&passing&ns=${NS}"
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
[ "$status" -eq 0 ]
echo "$output" | jq --raw-output '. | length'
}
wan federation via mesh gateways (#6884) 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.
2020-03-09 20:59:02 +00:00
function assert_alive_wan_member_count {
local EXPECT_COUNT=$1
run retry_long assert_alive_wan_member_count_once $EXPECT_COUNT
wan federation via mesh gateways (#6884) 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.
2020-03-09 20:59:02 +00:00
[ "$status" -eq 0 ]
}
function assert_alive_wan_member_count_once {
local EXPECT_COUNT=$1
GOT_COUNT=$(get_alive_wan_member_count)
[ "$GOT_COUNT" -eq "$EXPECT_COUNT" ]
}
function get_alive_wan_member_count {
run retry_default curl -sL -f "127.0.0.1:8500/v1/agent/members?wan=1"
[ "$status" -eq 0 ]
# echo "$output" >&3
echo "$output" | jq '.[] | select(.Status == 1) | .Name' | wc -l
}
function assert_service_has_healthy_instances_once {
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
local SERVICE_NAME=$1
local EXPECT_COUNT=$2
connect: fix failover through a mesh gateway to a remote datacenter (#6259) 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).
2019-08-05 18:30:35 +00:00
local DC=${3:-primary}
local NS=$4
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
GOT_COUNT=$(get_healthy_service_count "$SERVICE_NAME" "$DC" "$NS")
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
[ "$GOT_COUNT" -eq $EXPECT_COUNT ]
}
function assert_service_has_healthy_instances {
local SERVICE_NAME=$1
local EXPECT_COUNT=$2
connect: fix failover through a mesh gateway to a remote datacenter (#6259) 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).
2019-08-05 18:30:35 +00:00
local DC=${3:-primary}
local NS=$4
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
run retry_long assert_service_has_healthy_instances_once "$SERVICE_NAME" "$EXPECT_COUNT" "$DC" "$NS"
test: for envoy integration tests, wait until 's2' is healthy in consul before interrogating envoy (#6108) When the envoy healthy panic threshold was explicitly disabled as part of L7 traffic management it changed how envoy decided to load balance to endpoints in a cluster. This only matters when envoy is in "panic mode" aka "when you have a bunch of unhealthy endpoints". Panic mode sends traffic to unhealthy instances in certain circumstances. Note: Prior to explicitly disabling the healthy panic threshold, the default value is 50%. What was happening is that the test harness was bringing up consul the sidecars, and the service instances all at once and sometimes the proxies wouldn't have time to be checked by consul to be labeled as 'passing' in the catalog before a round of EDS happened. The xDS server in consul effectively queries /v1/health/connect/s2 and gets 1 result, but that one result has a 'critical' check so the xDS server sends back that endpoint labeled as UNHEALTHY. Envoy sees that 100% of the endpoints in the cluster are unhealthy and would enter panic mode and still send traffic to s2. This is why the test suites PRIOR to disabling the healthy panic threshold worked. They were _incorrectly_ passing. When the healthy panic threshol is disabled, envoy never enters panic mode in this situation and thus the cluster has zero healthy endpoints so load balancing goes nowhere and the tests fail. Why does this only affect the test suites for envoy 1.8.0? My guess is that https://github.com/envoyproxy/envoy/pull/4442 was merged into the 1.9.x series and somehow that plays a role. This PR modifies the bats scripts to explicitly wait until the upstream sidecar is healthy as measured by /v1/health/connect/s2?passing BEFORE trying to interrogate envoy which should make the tests less racy.
2019-07-10 20:58:25 +00:00
[ "$status" -eq 0 ]
}
function check_intention {
local SOURCE=$1
local DESTINATION=$2
curl -s -f "localhost:8500/v1/connect/intentions/check?source=${SOURCE}&destination=${DESTINATION}" | jq ".Allowed"
}
function assert_intention_allowed {
local SOURCE=$1
local DESTINATION=$2
run check_intention "${SOURCE}" "${DESTINATION}"
[ "$status" -eq 0 ]
[ "$output" = "true" ]
}
function assert_intention_denied {
local SOURCE=$1
local DESTINATION=$2
run check_intention "${SOURCE}" "${DESTINATION}"
[ "$status" -eq 0 ]
[ "$output" = "false" ]
}
function docker_consul {
local DC=$1
shift 1
docker run -i --rm --network container:envoy_consul-${DC}_1 consul-dev "$@"
}
function docker_consul_for_proxy_bootstrap {
local DC=$1
shift 1
docker run -i --rm --network container:envoy_consul-${DC}_1 consul-dev "$@"
}
function docker_wget {
local DC=$1
shift 1
docker run --rm --network container:envoy_consul-${DC}_1 docker.mirror.hashicorp.services/alpine:3.9 wget "$@"
}
function docker_curl {
local DC=$1
shift 1
docker run --rm --network container:envoy_consul-${DC}_1 --entrypoint curl consul-dev "$@"
}
function docker_exec {
if ! docker exec -i "$@"
then
echo "Failed to execute: docker exec -i $@" 1>&2
return 1
fi
}
function docker_consul_exec {
local DC=$1
shift 1
docker_exec envoy_consul-${DC}_1 "$@"
}
function kill_envoy {
local BOOTSTRAP_NAME=$1
local DC=${2:-primary}
pkill -TERM -f "envoy -c /workdir/$DC/envoy/${BOOTSTRAP_NAME}-bootstrap.json"
}
function must_match_in_statsd_logs {
local DC=${2:-primary}
run cat /workdir/${DC}/statsd/statsd.log
echo "$output"
COUNT=$( echo "$output" | grep -Ec $1 )
echo "COUNT of '$1' matches: $COUNT"
[ "$status" == 0 ]
[ "$COUNT" -gt "0" ]
}
function must_match_in_prometheus_response {
run curl -f -s $1/metrics
COUNT=$( echo "$output" | grep -Ec $2 )
echo "OUTPUT head -n 10"
echo "$output" | head -n 10
echo "COUNT of '$2' matches: $COUNT"
[ "$status" == 0 ]
[ "$COUNT" -gt "0" ]
}
function must_match_in_stats_proxy_response {
run curl -f -s $1/$2
COUNT=$( echo "$output" | grep -Ec $3 )
echo "OUTPUT head -n 10"
echo "$output" | head -n 10
echo "COUNT of '$3' matches: $COUNT"
[ "$status" == 0 ]
[ "$COUNT" -gt "0" ]
}
# must_fail_tcp_connection checks that a request made through an upstream fails,
# probably due to authz being denied if all other tests passed already. Although
# we are using curl, this only works as expected for TCP upstreams as we are
# checking TCP-level errors. HTTP upstreams will return a valid 503 generated by
# Envoy rather than a connection-level error.
function must_fail_tcp_connection {
# Attempt to curl through upstream
run curl --no-keepalive -s -v -f -d hello $1
echo "OUTPUT $output"
# Should fail during handshake and return "got nothing" error
[ "$status" == "52" ]
# Verbose output should enclude empty reply
echo "$output" | grep 'Empty reply from server'
}
function must_pass_tcp_connection {
run curl --no-keepalive -s -f -d hello $1
echo "OUTPUT $output"
[ "$status" == "0" ]
[ "$output" = "hello" ]
}
# must_fail_http_connection see must_fail_tcp_connection but this expects Envoy
# to generate a 503 response since the upstreams have refused connection.
function must_fail_http_connection {
# Attempt to curl through upstream
run curl --no-keepalive -s -i -d hello "$1"
echo "OUTPUT $output"
[ "$status" == "0" ]
local expect_response="${2:-403 Forbidden}"
# Should fail request with 503
echo "$output" | grep "${expect_response}"
}
# must_pass_http_request allows you to craft a specific http request to assert
# that envoy will NOT reject the request. Primarily of use for testing L7
# intentions.
function must_pass_http_request {
local METHOD=$1
local URL=$2
local DEBUG_HEADER_VALUE="${3:-""}"
local extra_args
if [[ -n "${DEBUG_HEADER_VALUE}" ]]; then
extra_args="-H x-test-debug:${DEBUG_HEADER_VALUE}"
fi
case "$METHOD" in
GET)
;;
DELETE)
extra_args="$extra_args -X${METHOD}"
;;
PUT|POST)
extra_args="$extra_args -d'{}' -X${METHOD}"
;;
*)
return 1
;;
esac
run curl --no-keepalive -v -s -f $extra_args "$URL"
[ "$status" == 0 ]
}
# must_fail_http_request allows you to craft a specific http request to assert
# that envoy will reject the request. Primarily of use for testing L7
# intentions.
function must_fail_http_request {
local METHOD=$1
local URL=$2
local DEBUG_HEADER_VALUE="${3:-""}"
local extra_args
if [[ -n "${DEBUG_HEADER_VALUE}" ]]; then
extra_args="-H x-test-debug:${DEBUG_HEADER_VALUE}"
fi
case "$METHOD" in
HEAD)
extra_args="$extra_args -I"
;;
GET)
;;
DELETE)
extra_args="$extra_args -X${METHOD}"
;;
PUT|POST)
extra_args="$extra_args -d'{}' -X${METHOD}"
;;
*)
return 1
;;
esac
# Attempt to curl through upstream
run curl --no-keepalive -s -i $extra_args "$URL"
echo "OUTPUT $output"
echo "$output" | grep "403 Forbidden"
}
function gen_envoy_bootstrap {
SERVICE=$1
ADMIN_PORT=$2
DC=${3:-primary}
IS_GW=${4:-0}
EXTRA_ENVOY_BS_ARGS="${5-}"
PROXY_ID="$SERVICE"
if ! is_set "$IS_GW"
then
PROXY_ID="$SERVICE-sidecar-proxy"
fi
if output=$(docker_consul_for_proxy_bootstrap "$DC" connect envoy -bootstrap \
-proxy-id $PROXY_ID \
-envoy-version "$ENVOY_VERSION" \
-admin-bind 0.0.0.0:$ADMIN_PORT ${EXTRA_ENVOY_BS_ARGS} 2>&1); then
# All OK, write config to file
echo "$output" > workdir/${DC}/envoy/$SERVICE-bootstrap.json
else
status=$?
# Command failed, instead of swallowing error (printed on stdout by docker
# it seems) by writing it to file, echo it
echo "$output"
return $status
fi
}
function read_config_entry {
local KIND=$1
local NAME=$2
local DC=${3:-primary}
docker_consul "$DC" config read -kind $KIND -name $NAME
}
function wait_for_namespace {
local NS="${1}"
local DC=${2:-primary}
retry_default docker_curl "$DC" -sLf "http://127.0.0.1:8500/v1/namespace/${NS}" >/dev/null
}
function wait_for_config_entry {
retry_default read_config_entry "$@" >/dev/null
}
function delete_config_entry {
local KIND=$1
local NAME=$2
retry_default curl -sL -XDELETE "http://127.0.0.1:8500/v1/config/${KIND}/${NAME}"
}
function register_services {
local DC=${1:-primary}
docker_consul_exec ${DC} sh -c "consul services register /workdir/${DC}/register/service_*.hcl"
}
function setup_upsert_l4_intention {
local SOURCE=$1
local DESTINATION=$2
local ACTION=$3
retry_default docker_curl primary -sL -XPUT "http://127.0.0.1:8500/v1/connect/intentions/exact?source=${SOURCE}&destination=${DESTINATION}" \
-d"{\"Action\": \"${ACTION}\"}" >/dev/null
}
function upsert_l4_intention {
local SOURCE=$1
local DESTINATION=$2
local ACTION=$3
retry_default curl -sL -XPUT "http://127.0.0.1:8500/v1/connect/intentions/exact?source=${SOURCE}&destination=${DESTINATION}" \
-d"{\"Action\": \"${ACTION}\"}" >/dev/null
}
function get_ca_root {
curl -s -f "http://localhost:8500/v1/connect/ca/roots" | jq -r ".Roots[0].RootCert"
}
function wait_for_agent_service_register {
local SERVICE_ID=$1
local DC=${2:-primary}
retry_default docker_curl "$DC" -sLf "http://127.0.0.1:8500/v1/agent/service/${SERVICE_ID}" >/dev/null
}
function set_ttl_check_state {
local CHECK_ID=$1
local CHECK_STATE=$2
local DC=${3:-primary}
case "$CHECK_STATE" in
pass)
;;
warn)
;;
fail)
;;
*)
echo "invalid ttl check state '${CHECK_STATE}'" >&2
return 1
esac
retry_default docker_curl "${DC}" -sL -XPUT "http://localhost:8500/v1/agent/check/warn/${CHECK_ID}"
}
function get_upstream_fortio_name {
local HOST=$1
local PORT=$2
local PREFIX=$3
local DEBUG_HEADER_VALUE="${4:-""}"
local extra_args
if [[ -n "${DEBUG_HEADER_VALUE}" ]]; then
extra_args="-H x-test-debug:${DEBUG_HEADER_VALUE}"
fi
# split proto if https:// is at the front of the host since the --resolve
# string needs just a bare host.
local PROTO=""
local CA_FILE=""
if [ "${HOST:0:8}" = "https://" ]; then
HOST="${HOST:8}"
PROTO="https://"
extra_args="${extra_args} --cacert /workdir/test-sds-server/certs/ca-root.crt"
fi
# We use --resolve instead of setting a Host header since we need the right
# name to be sent for SNI in some cases too.
run retry_default curl -v -s -f --resolve "${HOST}:${PORT}:127.0.0.1" $extra_args \
"${PROTO}${HOST}:${PORT}${PREFIX}/debug?env=dump"
# Useful Debugging but breaks the expectation that the value output is just
# the grep output when things don't fail
if [ "$status" != 0 ]; then
echo "GOT FORTIO OUTPUT: $output"
fi
[ "$status" == 0 ]
echo "$output" | grep -E "^FORTIO_NAME="
}
function assert_expected_fortio_name {
local EXPECT_NAME=$1
local HOST=${2:-"localhost"}
local PORT=${3:-5000}
local URL_PREFIX=${4:-""}
local DEBUG_HEADER_VALUE="${5:-""}"
run get_upstream_fortio_name ${HOST} ${PORT} "${URL_PREFIX}" "${DEBUG_HEADER_VALUE}"
echo "GOT: $output"
[ "$status" == 0 ]
[ "$output" == "FORTIO_NAME=${EXPECT_NAME}" ]
}
function assert_expected_fortio_name_pattern {
local EXPECT_NAME_PATTERN=$1
local HOST=${2:-"localhost"}
local PORT=${3:-5000}
local URL_PREFIX=${4:-""}
local DEBUG_HEADER_VALUE="${5:-""}"
GOT=$(get_upstream_fortio_name ${HOST} ${PORT} "${URL_PREFIX}" "${DEBUG_HEADER_VALUE}")
if [[ "$GOT" =~ $EXPECT_NAME_PATTERN ]]; then
:
else
echo "expected name pattern: $EXPECT_NAME_PATTERN, actual name: $GOT" 1>&2
return 1
fi
}
2020-12-16 07:09:55 +00:00
function get_upstream_fortio_host_header {
local HOST=$1
local PORT=$2
local PREFIX=$3
local DEBUG_HEADER_VALUE="${4:-""}"
local extra_args
if [[ -n "${DEBUG_HEADER_VALUE}" ]]; then
extra_args="-H x-test-debug:${DEBUG_HEADER_VALUE}"
fi
run retry_default curl -v -s -f -H"Host: ${HOST}" $extra_args \
"localhost:${PORT}${PREFIX}/debug"
[ "$status" == 0 ]
echo "$output" | grep -E "^Host: "
}
function assert_expected_fortio_host_header {
local EXPECT_HOST=$1
local HOST=${2:-"localhost"}
local PORT=${3:-5000}
local URL_PREFIX=${4:-""}
local DEBUG_HEADER_VALUE="${5:-""}"
GOT=$(get_upstream_fortio_host_header ${HOST} ${PORT} "${URL_PREFIX}" "${DEBUG_HEADER_VALUE}")
if [ "$GOT" != "Host: ${EXPECT_HOST}" ]; then
echo "expected Host header: $EXPECT_HOST, actual Host header: $GOT" 1>&2
return 1
fi
}