Fix#5812
LLDP conf Jinja2 Template does not verify IPv4 address and can use IPv6 version. This issue does not effect control LLDP daemon. Issue can be reproduced via `test_snmp_lldp` test. LLDP conf Jinja2 Template selects first item from the list of mgmt interfaces.
TESTBED_1 LLDP conf
```
configure ports eth0 lldp portidsubtype local eth0
configure system ip management pattern FC00:3::32
configure system hostname dut-1
```
TESTBED_2 LLDP conf
```
configure ports eth0 lldp portidsubtype local eth0
configure system ip management pattern 10.22.24.61
configure system hostname dut-2
```
TESTBED_1 MGMT_INTERFACE
```
$ redis-cli -n 4 keys "*" | grep MGMT_INTERFACE
MGMT_INTERFACE|eth0|10.22.24.53/23
MGMT_INTERFACE|eth0|FC00:3::32/64
```
TESTBED_2 MGMT_INTERFACE
```
$ redis-cli -n 4 keys "*" | grep MGMT_INTERFACE
MGMT_INTERFACE|eth0|FC00:3::32/64
MGMT_INTERFACE|eth0|10.22.24.61/23
```
Signed-off-by: Petro Bratash <petrox.bratash@intel.com>
* Fix for LLDP advertisments being sent with wrong information.
Since lldpd is starting before lldpmgr, some advertisment packets might sent with default value, mac address as Port ID.
This fix hold the packets from being sent by the lldpd until all interfaces are well configured by the lldpmgrd.
Signed-off-by: Shlomi Bitton <shlomibi@nvidia.com>
* Fix comments
* Fix unit-test output caused a failure during build
* Add 'run_cmd' function and use it
* Resume lldpd even if port init timeout reached
* We don't need configure anything until we have interfaces created
* Don't run lldpcli for a port, until a port is up and running
* Remove lldpd socket before starting lldpd
* Fix sample files for lldpd configuration
* Another attempt to make the test working
* Quick fix for lldpd paused after start bug
Current lldp.conf.j2 template demands the presence of MGMT_INTERFACE attribute in configDB, and by extension, also in config_db.json file. However, MGMT_INTERFACE configuration attribute is optional, so lldp shouldn't bail out if this one isn't provided in configuration.
For this reason, no lldp.conf file is ever created in lldp's container, and lldpd defaults to advertise the mac-address of the connected interface, instead of the interface name.
The fix is to simply relax this jinja2 statement to verify if MGMT_INTERFACE attribute is present.
The current DEVICE_NEIGHBOR format doesn't support multiple different
ports connecting with same neighbor. Thus the lldpd.conf file is not
generated correctly, causing missing information for LAG members.
This fix reverts the data structure in the previous version of
minigraph parser - using local port as the key and remote port/device
as the value of the map. Sample format is:
DEVICE_NEIGHBOR['Ethernet124'] = {
'name': 'ARISTA04T1',
'port': 'Ethernet1/1'
}
The corresponding unit test in test_cfggen is updated.
Add one more unit test for lldpd.conf.j2 verification.
Signed-off-by: Shu0T1an ChenG <shuche@microsoft.com>
Modify minigraph parser output format so it fit DB schema
Modify configuration templates to fit new schema
Systemd services dependencies are modified so database starts before any configuration consumer
Add a sonic-config-engine to help generate config file based on minigraph and other data on runtime. Modify fpm, teamd, lldp, snmp, and platform-monitor docker to use sonic-config-engine to generate config in docker upon load.
