BGP Control Plane Operation Guide
This document provides guidance on how to operate the BGP Control Plane.
BGP Cilium CLI
Installation
Install the latest version of the Cilium CLI. The Cilium CLI can be used to install Cilium, inspect the state of a Cilium installation, and enable/disable various features (e.g. clustermesh, Hubble).
CILIUM_CLI_VERSION=$(curl -s https://raw.githubusercontent.com/cilium/cilium-cli/main/stable.txt)
CLI_ARCH=amd64
if [ "$(uname -m)" = "aarch64" ]; then CLI_ARCH=arm64; fi
curl -L --fail --remote-name-all https://github.com/cilium/cilium-cli/releases/download/${CILIUM_CLI_VERSION}/cilium-linux-${CLI_ARCH}.tar.gz{,.sha256sum}
sha256sum --check cilium-linux-${CLI_ARCH}.tar.gz.sha256sum
sudo tar xzvfC cilium-linux-${CLI_ARCH}.tar.gz /usr/local/bin
rm cilium-linux-${CLI_ARCH}.tar.gz{,.sha256sum}
CILIUM_CLI_VERSION=$(curl -s https://raw.githubusercontent.com/cilium/cilium-cli/main/stable.txt)
CLI_ARCH=amd64
if [ "$(uname -m)" = "arm64" ]; then CLI_ARCH=arm64; fi
curl -L --fail --remote-name-all https://github.com/cilium/cilium-cli/releases/download/${CILIUM_CLI_VERSION}/cilium-darwin-${CLI_ARCH}.tar.gz{,.sha256sum}
shasum -a 256 -c cilium-darwin-${CLI_ARCH}.tar.gz.sha256sum
sudo tar xzvfC cilium-darwin-${CLI_ARCH}.tar.gz /usr/local/bin
rm cilium-darwin-${CLI_ARCH}.tar.gz{,.sha256sum}
See the full page of releases.
Clone the Cilium GitHub repository so that the Cilium CLI can access the latest unreleased Helm chart from the main branch:
git clone git@github.com:cilium/cilium.git
cd cilium
Cilium BGP state can be inspected via cilium bgp
subcommand.
# cilium bgp --help
Access to BGP control plane
Usage:
cilium bgp [command]
Available Commands:
peers Lists BGP peering state
routes Lists BGP routes
Flags:
-h, --help help for bgp
Global Flags:
--context string Kubernetes configuration context
--helm-release-name string Helm release name (default "cilium")
--kubeconfig string Path to the kubeconfig file
-n, --namespace string Namespace Cilium is running in (default "kube-system")
Use "cilium bgp [command] --help" for more information about a command.
Peers
cilium bgp peers
command displays current peering states from all nodes in the kubernetes
cluster.
In the following example, peering status is displayed for two nodes in the cluster.
# cilium bgp peers
Node Local AS Peer AS Peer Address Session State Uptime Family Received Advertised
bgpv2-cplane-dev-service-control-plane 65001 65000 fd00:10::1 established 33m26s ipv4/unicast 2 2
ipv6/unicast 2 2
bgpv2-cplane-dev-service-worker 65001 65000 fd00:10::1 established 33m25s ipv4/unicast 2 2
ipv6/unicast 2 2
Using this command, you can validate BGP session state is established
and expected number
of routes are being advertised to the peers.
Routes
cilium bgp routes
command displays detailed information about local BGP routing table and per peer
advertised routing information.
In the following example, the local BGP routing table for IPv4/Unicast address family is shown for two nodes in the cluster.
# cilium bgp routes available ipv4 unicast
Node VRouter Prefix NextHop Age Attrs
bgpv2-cplane-dev-service-control-plane 65001 10.1.0.0/24 0.0.0.0 46m45s [{Origin: i} {Nexthop: 0.0.0.0}]
bgpv2-cplane-dev-service-worker 65001 10.1.1.0/24 0.0.0.0 46m45s [{Origin: i} {Nexthop: 0.0.0.0}]
Similarly, you can inspect per peer advertisements using following command.
# cilium bgp routes advertised ipv4 unicast
Node VRouter Peer Prefix NextHop Age Attrs
bgpv2-cplane-dev-service-control-plane 65001 fd00:10::1 10.1.0.0/24 fd00:10:0:1::2 47m0s [{Origin: i} {AsPath: 65001} {Communities: 65000:99} {MpReach(ipv4-unicast): {Nexthop: fd00:10:0:1::2, NLRIs: [10.1.0.0/24]}}]
bgpv2-cplane-dev-service-worker 65001 fd00:10::1 10.1.1.0/24 fd00:10:0:2::2 47m0s [{Origin: i} {AsPath: 65001} {Communities: 65000:99} {MpReach(ipv4-unicast): {Nexthop: fd00:10:0:2::2, NLRIs: [10.1.1.0/24]}}]
You can validate the BGP attributes are advertised based on configured CiliumBGPAdvertisement resources.
Policies
Cilium BGP installs GoBGP policies for managing per peer advertisement and BGP attributes. As this is an internal implementation detail, it is not exposed via Cilium CLI. However, for debugging purpose you can inspect installed BGP policies using cilium-dbg CLI from the Cilium agent pod.
/home/cilium# cilium-dbg bgp route-policies
VRouter Policy Name Type Match Peers Match Prefixes (Min..Max Len) RIB Action Path Actions
65001 65000-ipv4-PodCIDR export fd00:10::1/128 10.1.0.0/24 (24..24) accept AddCommunities: [65000:99]
65001 65000-ipv6-PodCIDR export fd00:10::1/128 fd00:10:1::/64 (64..64) accept AddCommunities: [65000:99]
65001 allow-local import accept
CiliumBGPClusterConfig Status
CiliumBGPClusterConfig may report some configuration errors in the
.status.conditions
caught at runtime. Currently, the following conditions
are defined.
Condition Name |
Description |
---|---|
|
|
|
The PeerConfig specified in the |
|
There is an another CiliumBGPClusterConfig selecting the same node. |
CiliumBGPPeerConfig Status
CiliumBGPPeerConfig may report some configuration errors in the
.status.conditions
caught at runtime. Currently, the following conditions
are defined.
Condition Name |
Description |
---|---|
|
The Secret specified in the |
CiliumBGPNodeConfig Status
Each Cilium node on which BGP control plane is enabled based on CiliumBGPClusterConfig
node selector gets associated
CiliumBGPNodeConfig
resource. CiliumBGPNodeConfig
resource is the source of BGP configuration for the
node, it is managed by Cilium operator.
Status field of CiliumBGPNodeConfig
maintains real-time BGP operational state. This can be used for
automation or monitoring purposes.
In the following example, you can see BGP instance state from node bgpv2-cplane-dev-service-worker
.
# kubectl describe ciliumbgpnodeconfigs bgpv2-cplane-dev-service-worker
Name: bgpv2-cplane-dev-service-worker
Namespace:
Labels: <none>
Annotations: <none>
API Version: cilium.io/v2alpha1
Kind: CiliumBGPNodeConfig
Metadata:
Creation Timestamp: 2024-10-17T13:59:44Z
Generation: 1
Owner References:
API Version: cilium.io/v2alpha1
Kind: CiliumBGPClusterConfig
Name: cilium-bgp
UID: f0c23da8-e5ca-40d7-8c94-91699cf1e03a
Resource Version: 1385
UID: fc88be94-37e9-498a-b9f7-a52684090d80
Spec:
Bgp Instances:
Local ASN: 65001
Name: 65001
Peers:
Name: 65000
Peer ASN: 65000
Peer Address: fd00:10::1
Peer Config Ref:
Group: cilium.io
Kind: CiliumBGPPeerConfig
Name: cilium-peer
Status:
Bgp Instances:
Local ASN: 65001
Name: 65001
Peers:
Established Time: 2024-10-17T13:59:50Z
Name: 65000
Peer ASN: 65000
Peer Address: fd00:10::1
Peering State: established
Route Count:
Advertised: 2
Afi: ipv4
Received: 2
Safi: unicast
Advertised: 2
Afi: ipv6
Received: 2
Safi: unicast
Timers:
Applied Hold Time Seconds: 90
Applied Keepalive Seconds: 30
Events: <none>
Disabling CRD Status Report
CRD status reporting is useful for troubleshooting, making it useful to enable
in general. However, for large clusters with a lot of nodes or BGP policies,
CRD status reporting may add a significant API server load. To disable status
reporting, set the bgpControlPlane.statusReport.enabled
Helm value to
false
. Doing so disables status reporting and clears the currently reported
status.
Logs
BGP Control Plane logs can be found in the Cilium operator (only for BGPv2) and the Cilium agent logs.
The operator logs are tagged with subsys=bgp-cp-operator
. You can use this tag to filter
the logs as in the following example:
kubectl -n kube-system logs <cilium operator pod name> | grep "subsys=bgp-cp-operator"
The agent logs are tagged with subsys=bgp-control-plane
. You can use this tag to filter
the logs as in the following example:
kubectl -n kube-system logs <cilium agent pod name> | grep "subsys=bgp-control-plane"
Metrics
Metrics exposed by BGP Control Plane are listed in the metrics document.
Restarting an Agent
When you restart the Cilium agent, the BGP session will be lost because the BGP speaker is integrated within the Cilium agent. The BGP session will be restored once the Cilium agent is restarted. However, while the Cilium agent is down, the advertised routes will be removed from the BGP peer. As a result, you may temporarily lose connectivity to the Pods or Services. You can enable the Graceful Restart to continue forwarding traffic to the Pods or Services during the agent restart.
Upgrading or Downgrading Cilium
When you upgrade or downgrade Cilium, you must restart the Cilium agent. For more details about the agent restart, see Restarting an Agent section.
Note that with BGP Control Plane, it’s especially important to pre-pull the
agent image by following the preflight process before
upgrading Cilium. Image pull is time-consuming and error-prone because it
involves network communication. If the image pull takes longer, it may exceed
the Graceful Restart time (restartTimeSeconds
) and cause the BGP peer to
withdraw routes.
Shutting Down a Node
When you need to shut down a node for maintenance, you can follow the steps below to avoid packet loss as much as possible.
Drain the node to evict all workloads. This will remove all Pods on the node from the Service endpoints and prevent Services with
externalTrafficPolicy=Cluster
from redirecting traffic to the node.kubectl drain <node-name> --ignore-daemonsets
Reconfigure the BGP sessions by modifying or removing the CiliumBGPPeeringPolicy or CiliumBGPClusterConfig node selector label on the Node object. This will shut down all BGP sessions on the node.
# Assuming you select the node by the label enable-bgp=true kubectl label node <node-name> --overwrite enable-bgp=false
Wait for a while until the BGP peer removes routes towards the node. During this period, the BGP peer may still send traffic to the node. If you shut down the node without waiting for the BGP peer to remove routes, it will break the ongoing traffic of
externalTrafficPolicy=Cluster
Services.Shut down the node.
In step 3, you may not be able to check the peer status and may want to wait for a specific period of time without checking the actual peer status. In this case, you can roughly estimate the time like the following:
If you disable the BGP Graceful Restart feature, the BGP peer should withdraw routes immediately after step 2.
If you enable the BGP Graceful Restart feature, there are two possible cases.
If the BGP peer supports the Graceful Restart with Notification (RFC 8538), it will withdraw routes after the Stale Timer (defined in the RFC 8538#section-4.1) expires.
If the BGP peer does not support the Graceful Restart with Notification, it will withdraw routes immediately after step 2 because the BGP Control Plane sends the BGP Notification to the peer when you unselect the node.
The above estimation is a theoretical value, and the actual time always depends on the BGP peer’s implementation. Ideally, you should check the peer router’s actual behavior in advance with your network administrator.
Warning
Even if you follow the above steps, some ongoing Service traffic originally destined for the node may be reset because, after the route withdrawal and ECMP rehashing, the traffic is redirected to a different node, and the new node may select a different endpoint.
Failure Scenarios
This document describes common failure scenarios that you may encounter when using the BGP Control Plane and provides guidance on how to mitigate them.
Cilium Agent Down
If the Cilium agent goes down, the BGP session will be lost because the BGP speaker is integrated within the Cilium agent. The BGP session will be restored once the Cilium agent is restarted. However, while the Cilium agent is down, the advertised routes will be removed from the BGP peer. As a result, you may temporarily lose connectivity to the Pods or Services.
Mitigation
The recommended way to address this issue is by enabling the Graceful Restart feature. This feature allows the BGP peer to retain routes for a specific period of time after the BGP session is lost. Since the datapath remains active even when the agent is down, this will prevent the loss of connectivity to the Pods or Services.
When you can’t use BGP Graceful Restart, you can take the following actions, depending on the kind of routes you are using:
PodCIDR routes
If you are advertising PodCIDR routes, pods on the failed node will be unreachable from the external network. If the failure only occurs on a subset of the nodes in the cluster, you can drain the unhealthy nodes to migrate the pods to other nodes.
Service routes
If you are advertising service routes, the load balancer (KubeProxy or Cilium KubeProxyReplacement) may become unreachable from the external network. Additionally, ongoing connections may be redirected to different nodes due to ECMP rehashing on the upstream routers. When the load balancer encounters unknown traffic, it will select a new endpoint. Depending on the load balancer’s backend selection algorithm, the traffic may be directed to a different endpoint than before, potentially causing the connection to be reset.
If your upstream routers support ECMP with Resilient Hashing, enabling
it may help to keep the ongoing connections forwarded to the same node.
Enabling the Maglev Consistent Hashing feature in Cilium may also help since it increases
the probability that all nodes select the same endpoint for the same flow.
However, it only works for the externalTrafficPolicy: Cluster
. If the
Service’s externalTrafficPolicy
is set to Local
, it is inevitable that
all ongoing connections with the endpoints on the failed node, and connections
forwarded to a different node than before, will be reset.
Node Down
If the node goes down, the BGP sessions from this node will be lost. The peer
will withdraw the routes advertised by the node immediately or takes some time
to stop forwarding traffic to the node depending on the Graceful Restart settings.
The latter case is problematic when you advertise the route to a Service with
externalTrafficPolicy=Cluster
because the peer will continue to forward traffic
to the unavailable node until the restart timer (which is 120s by default) expires.
Mitigation
Involuntary Shutdown
When a node is involuntarily shut down, there’s no direct mitigation. You can choose to not use the BGP Graceful Restart feature, depending on the trade-off between the failure detection time vs stability provided by graceful restart in cases of Cilium pod restarts.
Disabling the Graceful Restart allows the BGP peer to withdraw routes faster. Even if the node is shut down without BGP Notification or TCP connection close, the worst case time for peer to withdraw routes is the BGP hold time. When the Graceful Restart is enabled, the BGP peer may need hold time + restart time to withdraw routes received from the node.
Voluntary Shutdown
When you voluntarily shut down a node, you can follow the steps described in the Shutting Down a Node section to avoid packet loss as much as possible.
Peering Link Down
If the peering link between the BGP peers goes down, usually, both the BGP session and datapath connectivity will be lost. However, there may be a period during which the datapath connectivity is lost while the BGP session remains up and routes are still being advertised. This can cause the BGP peer to send traffic over the failed link, resulting in dropped packets. The length of this period depends on which link is down and the BGP configuration.
If the link directly connected to the Node goes down, the BGP session will likely be lost immediately because the Linux kernel detects the link failure and shuts down the TCP session right away. If a link not directly connected to the Node goes down, the BGP session will be lost after the hold timer expires, which is set to 90 seconds by default.
Mitigation
To make link detection failure fast, you can adjust holdTimeSeconds
and
keepAliveTimeSeconds
in the BGP configuration to the shorter value.
However, the minimal possible values are holdTimeSeconds=3
and
keepAliveTimeSeconds=1
. The general approach to make failure detection faster is to
use BFD (Bidirectional Forwarding Detection), but currently, Cilium does not
support it.
Cilium Operator Down
The Cilium operator is responsible for translating CiliumBGPClusterConfig
to
the per node CiliumBGPNodeConfig
resource. If the Cilium operator is down,
provisioning of BGP control plane will be stopped.
Similarly, PodCIDR allocation by IPAM, and LoadBalancer IP allocation by LB-IPAM are stopped. Therefore, the advertisement of new and withdrawal of old PodCIDR and Service VIP routes will be stopped as well.
Mitigation
There’s no direct mitigation in terms of the BGP. However, running the Cilium Operator with a high-availability setup will make the Cilium Operator more resilient to failures.
Service Losing All Backends
If all service backends are gone due to an outage or a configuration mistake, BGP
Control Plane behaves differently depending on the Service’s
externalTrafficPolicy
. When the externalTrafficPolicy
is set to
Cluster
, the Service’s VIP remains advertised from all nodes selected by the
CiliumBGPPeeringPolicy
or CiliumBGPClusterConfig
. When the externalTrafficPolicy
is set to Local
, the advertisement stops entirely because the Service’s VIP is only advertised
from the node where the Service backends are running.
Mitigation
There’s no direct mitigation in terms of the BGP. In general, you should prevent the Service backends from being all gone by Kubernetes features like PodDisruptionBudget.