Skip to content

Commit

Permalink
Update pkg/crds/calico/crd.projectcalico.org_felixconfigurations.yaml (
Browse files Browse the repository at this point in the history
  • Loading branch information
rene-dekker authored Jan 10, 2025
1 parent ecdc1ce commit 24c6261
Showing 1 changed file with 81 additions and 2 deletions.
83 changes: 81 additions & 2 deletions pkg/crds/calico/crd.projectcalico.org_felixconfigurations.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -107,6 +107,80 @@ spec:
- Userspace
- BPFProgram
type: string
bpfConntrackTimeouts:
description: |-
BPFConntrackTimers overrides the default values for the specified conntrack timer if
set. Each value can be either a duration or `Auto` to pick the value from
a Linux conntrack timeout.
Configurable timers are: CreationGracePeriod, TCPSynSent,
TCPEstablished, TCPFinsSeen, TCPResetSeen, UDPTimeout, GenericTimeout,
ICMPTimeout.
Unset values are replaced by the default values with a warning log for
incorrect values.
properties:
creationGracePeriod:
description: |2-
CreationGracePeriod gives a generic grace period to new connection
before they are considered for cleanup [Default: 10s].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
genericTimeout:
description: |-
GenericTimeout controls how long it takes before considering this
entry for cleanup after the connection became idle. If set to 'Auto', the
value from nf_conntrack_generic_timeout is used. If nil, Calico uses its
own default value. [Default: 10m].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
icmpTimeout:
description: |-
ICMPTimeout controls how long it takes before considering this
entry for cleanup after the connection became idle. If set to 'Auto', the
value from nf_conntrack_icmp_timeout is used. If nil, Calico uses its
own default value. [Default: 5s].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
tcpEstablished:
description: |-
TCPEstablished controls how long it takes before considering this entry for
cleanup after the connection became idle. If set to 'Auto', the
value from nf_conntrack_tcp_timeout_established is used. If nil, Calico uses
its own default value. [Default: 1h].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
tcpFinsSeen:
description: |-
TCPFinsSeen controls how long it takes before considering this entry for
cleanup after the connection was closed gracefully. If set to 'Auto', the
value from nf_conntrack_tcp_timeout_time_wait is used. If nil, Calico uses
its own default value. [Default: Auto].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
tcpResetSeen:
description: |-
TCPFinsSeen controls how long it takes before considering this entry for
cleanup after the connection was aborted. If nil, Calico uses its own
default value. [Default: 40s].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
tcpSynSent:
description: |-
TCPSynSent controls how long it takes before considering this entry for
cleanup after the last SYN without a response. If set to 'Auto', the
value from nf_conntrack_tcp_timeout_syn_sent is used. If nil, Calico uses
its own default value. [Default: 20s].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
udpTimeout:
description: |-
UDPTimeout controls how long it takes before considering this entry for
cleanup after the connection became idle. If nil, Calico uses its own
default value. [Default: 60s].
pattern: ^(([0-9]*(\.[0-9]*)?(ms|s|h|m|us)+)+|Auto)$
type: string
type: object
bpfDSROptoutCIDRs:
description: |-
BPFDSROptoutCIDRs is a list of CIDRs which are excluded from DSR. That is, clients
Expand Down Expand Up @@ -1055,8 +1129,13 @@ spec:
to use for the Wireguard routing rule. [Default: 99]'
type: integer
wireguardThreadingEnabled:
description: 'WireguardThreadingEnabled controls whether Wireguard
has NAPI threading enabled. [Default: false]'
description: |-
WireguardThreadingEnabled controls whether Wireguard has Threaded NAPI enabled. [Default: false]
This increases the maximum number of packets a Wireguard interface can process.
Consider threaded NAPI only if you have high packets per second workloads that are causing dropping packets due to a saturated `softirq` CPU core.
There is a [known issue](https://lore.kernel.org/netdev/CALrw=nEoT2emQ0OAYCjM1d_6Xe_kNLSZ6dhjb5FxrLFYh4kozA@mail.gmail.com/T/) with this setting
that may cause NAPI to get stuck holding the global `rtnl_mutex` when a peer is removed.
Workaround: Make sure your Linux kernel [includes this patch](https://github.com/torvalds/linux/commit/56364c910691f6d10ba88c964c9041b9ab777bd6) to unwedge NAPI.
type: boolean
workloadSourceSpoofing:
description: |-
Expand Down

0 comments on commit 24c6261

Please sign in to comment.