| Age | Commit message (Collapse) | Author |
|---|
|
VPP: T1797: Optimizations for VPP memory allocation during startup
|
|
- changed memory requirement from total to available. This allows to start on
systems with less total memory and protects from startup on systems with
overloaded memory.
- prevent startup if sysctl settings were not applied. This protects from
situations when the system cannot allocate enough hugepages or apply other
sysctl settings.
|
|
without vrf
|
|
* set protocols ospf aggregation timer <seconds>
* set protocols ospf summary-address x.x.x.x/x [tag 1-4294967295]
* set protocols ospf summary-address x.x.x.x/x no-advertise
|
|
After an interface is added/removed from VPP, it will be reinitialized, which
allows reconfiguring IP addresses on it.
Also modified VPP load priority to start before interfaces, and avoid
reconfiguration during boot.
|
|
T1797: VPP verify minimal installed memory and apply sysctl
|
|
Do not allow configure VPP if on the systems with low amount
installed memory
Add sysctl VPP parameters (hugepages, kernel.shmmax)
|
|
|
|
T1797: Add initial vpp configuration
|
|
- added extra renaming operation to be sure that interface has the same name as
before in the system after it was moved from VPP to kernel
- added extra check after PCI device removal/adding
- added check for proper `retval` for CPI calls where it is available
- replaced empty return with an error in `_get_pci_address_by_interface()`
because not resolved address will lead to inconsistency of the system later
|
|
- added ability to add/remove interfaces without system reboot
- added `attempts` and `interval` to the VPP API connection. This is helpful in
case of high system load or when VPP was just started and API is not yet
available.
- added exceptions to API calls. This allows handling errors in communication
with API properly in conf-mode scripts.
- fixed PCI address search in VPP to match Linux kernel and ethtool style
- fixed systemd daemons control - first reload, then restart
- removed debug prints
- removed `vm.nr_hugepages` configuration. It is not required now but increases
RAM requirements a lot.
|
|
Use info from both ethtool and VPP to find PCI address for an
interface.
|
|
T5304: Container add volume bind propagation option
|
|
Replaced CLI commands with API calls.
CLI commands still can be used via:
```
vpp_control = VPPControl()
vpp_control.cli_cmd('command_here')
```
|
|
set container name c1 volume myvlm propagation rshared
|
|
Add initial configuration mode for VPP (PoC)
set vpp cpu corelist-workers '2'
set vpp cpu main-core '1'
set vpp interface eth1 num-rx-desc '256'
set vpp interface eth1 num-rx-queues '512'
set vpp interface eth1 num-tx-desc '256'
set vpp interface eth1 num-tx-queues '512'
set vpp interface eth1 pci '0000:02:00.0'
set vpp interface eth1 rx-mode 'polling'
set vpp interface eth2 pci '0000:08:00.0'
Limitation:
- 'set vpp interface ethX pci auto' works only per first
commit, then interface detached from default stack and creates
tun interface 'ethX' to communicate with default stack. In this
case we can't get PCI address via ethtool for 'tun' interfaces.
But we can set pci address manualy.
- Interface sync between default stack and VPP-DPDK stack
After vpp change it doesn't trigger iproute2 for changes
(should be written later)
I.e. if we change something in vpp per each commit it restarts
vpp.service it gets empty interface config as we don't configure vpp
directly and it should be configured via iproute2
But then if we do any change on interface (for example description)
it gets IP address, MTU, state, etc.
|
|
|
|
Check upper bound as defined in Debian Policy Manual. Without this
check, user 'nobody' will not be available.
|
|
|
|
|
|
... this is a step towards a new and better implementation that will utilize
VPP.
|
|
|
|
Prevent failure when the user enters a partial CLI command without any
address specified.
Also, apply some minor formatting changes.
|
|
T5260: Do not use deprecated python crypt module
|
|
Remove the quotes, this will always return the hash for string "password".
|
|
|
|
Templatize systemd override for ddclient service and move the generated
override files in /run. This ensures that the override files are always
generated afresh after boot.
Additionally, simplify the systemd override file by removing the
redundant/superfluous overrides.
|
|
|
|
|
|
ddclient implementation of dualstack for dyndns2 protocol is targeted
for dyn.com (dyndns.org) only. Dualstack won't work for other servers
supporting dyndns2 protocol (for example, dyn.dns.he.net).
|
|
Apply next round of configuration tree updates to 'service dns dynamic'
with the following changes:
- Migrate `service dns dynamic interface <interface> [use-web]`
to `service dns dynamic address <interface>`
or `service dns dynamic address web [web-options]`
This communicates the intent that dynamic dns IP address is detected
in only one way - using the `<interface>` or using an external web
request, not both.
- When using external web request, (`service dns dynamic address web`),
external url is optional (`web-options url`). Ddclient defaults are
used when unspecified,
- Rename all config `login` to `username` for consistency and also to
align better with alternative ddclient backends in consideration.
- Apply global 'ipv6-enable' to per service 'ip-version: ipv6'. Selecting
usage of IPv4 or IPv6 (or both simultaneously) is now at per service
(protocol) level instead of global level. This allows more control on
the ability to select IPv4 in some cases and IPv6 in some other cases
wherever supported by the underlying ddclient protocol.
- While the IP address (and by extension, the detection mechanism) is
global, the way it is applied to a particular ddclient protocol depends
on whether it supports IPv4 or IPv6 or both.
- Related to the above, this also prevents generating incorrect config
file (`ddclient.conf`) with multiple global sections leading to an
unpredictable behavior of ddclient.
- Implement provider (protocol) specific custom tweaks whenever possible
(e.g., `zone`, `username`, `server` are not necessary in all cases).
- Move service name from a combination of 'protocol' (with protocol
config autodetected) and custom (with protocol config specified) to a
single 'service' key. This allows for consisent setup of multiple
config for the same ddclient protocol (with different options and
credentials). This also avoid ambiguity with usual networking term
'protocol' and ddclient specific term 'protocol' (and can change with
a move to a different backend).
- Apply upfront XML constraints and validations consistently wherever
applicable.
- RFC2136 specific change: Rename rfc2136 config `record` to `host-name`
for consistency.
- Cloudflare specific change: While ddclient still supports authenticating
with email and global auth key, skipping `username` in config will
indicate the intent to use API token authentication (with special
'token' literal as `username`).
|
|
|
|
|
|
|
|
T5210:VPN:Fix typo in Warning
|
|
|
|
|
|
This is a radvd limitation.
|
|
|
|
To accept a Proxy Protocol header on incoming TCP connections,
add an accept-proxy parameter to the bind line in a frontend section.
This parameter detects both Proxy Protocol version 1 (text format)
and Proxy Protocol version 2 (binary format).
set load-balancing reverse-proxy backend <tag> server <tag> send-proxy
|
|
Improve and refactoring "load-balancing reverse-proxy"
- replace 'reverse-proxy server <tag>'
=> 'reverse-proxy service <tag>'
- replace 'reverse-proxy global-parameters tls <xxx>'
=> 'reverse-proxy global-parameters tls-version-min xxx'
=> 'reverse-proxy global-parameters ssl-bind-ciphers xxx'
- replace 'reverse-proxy service https rule <tag> set server 'xxx'
=> 'reverse-proxy service https rule <tag> set backend 'xxx'
'service https rule <tag> domain-name xxx' set as multinode
|
|
T5222: Add load-balancing for web traffic
|
|
|
|
Avoids the following warning, and any external service from
re-activating syslog via systemd socket activation:
frebib@vyos# commit
[ system syslog ]
DEBUG/COMMAND returned (err):
Warning: Stopping syslog.service, but it can still be activated by:
syslog.socket
DEBUG/COMMAND cmd 'systemctl stop syslog.service'
Signed-off-by: Joe Groocock <me@frebib.net>
|
|
os.unlink() is the correct function:
Traceback (most recent call last):
File "/usr/libexec/vyos/conf_mode/system-syslog.py", line 146, in <module>
generate(c)
File "/usr/libexec/vyos/conf_mode/system-syslog.py", line 114, in generate
os.path.unlink(rsyslog_conf)
^^^^^^^^^^^^^^
AttributeError: module 'posixpath' has no attribute 'unlink'
Signed-off-by: Joe Groocock <me@frebib.net>
|
|
There is no need for the backend code to call ethtool and try to change speed or
duplex settings every time there is a change in the interface configuration,
but no change for the speed/duplex subnodes. This also makes the commit itself
faster when working with ethernet interfaces.
Bonus: no repeating CLI messages that the driver does not support speed/duplex
changes, as we do not change anything here.
Extension to commit f2ecc9710 ("ethernet: T3891: honor auto-negotiation support
per NIC")
|
|
When enabling identity-based-config, users can add arbitrary config keys
that are processed by ocserv. The user "must know" what he is been doing, as
invalid config option will make the ocserv daemon go ... whoop!
Thus add a warning and inform the user about this setting.
|
|
ocserv: T3896: add CLI options to configure ocserv config-per-user/group
|
|
T5219: ddclient: Allow not set login for Cloudflare API token
|
|
|
