GitLab

netlab/:
+ cf-ospf-area/data/* added,
+ cf-ospf-authentication/data/* added,
+ cf-ospf-base/data/* added,
+ cf-ospf-bfd/data/* added,
+ cf-ospf-custom/data/* added,
+ cf-ospf-default/data/* added,
+ cf-ospf-nbma/data/* added,
+ cf-ospf-priority/data/* added,
+ cf-ospf-ptmp/data/* added,
+ cf-ospf-vrf/data/* added.

netlab/:
+ cf-ospf-area/test-cf-ospf-area.py updated,
+ cf-ospf-authentication/test-cf-ospf-authentication.py updated,
+ cf-ospf-base/test-cf-ospf-base.py updated,
+ cf-ospf-bfd/test-cf-ospf-bfd.py updated,
+ cf-ospf-custom/test-cf-ospf-custom.py updated,
+ cf-ospf-default/test-cf-ospf-default.py updated,
+ cf-ospf-nbma/test-cf-ospf-nbma.py updated,
+ cf-ospf-priority/test-cf-ospf-priority.py updated,
+ cf-ospf-ptmp/test-cf-ospf-ptmp.py updated,
+ cf-ospf-vrf/test-cf-ospf-vrf.py updated.

netlab/tests/strip_output:
+ function for filtering DTime column added.

netlab/tests/get_stdout_ospf_neighbors:
+ shell script for cleaning the saved tables added.

netlab/tests/kernel.py:
+ function for saving and checking tables added.

 - Remove unnecessary 'networks' section on non-ABRs
 - Switch stub interfaces on m3-m6 to areas 1,2
 - Aggregate 10.0.x.0/24 networks per area
 - Change 10.0.7.0/24 to 10.0.8.0/24 to fit aggregation
 - Switch area 2 to NSSA

IPv6 prefix may start with ::

bird-tools/netlab/:
+ cf-ospf-area/data/* added.

Test-case folder contains:
+ Bird configuration files,
+ Netlab configuration file,
+ Python test suite.

The workaround removes 'pref medium' from ECMP routes to match behavior
on older kernels/iputils. It should also handle host routes.

This was broken in recent configuration change, updated data were not
committed for cf-ospf-authentication.

That matches previous behavior of stop, so we need at least
temporarily to keep that as default.

Runtest should fail when stop command failed

Check for proper BIRD shutdown, handle stuck process or crash
during shutdown. Some changes done by commiter.

Netlab ignored most errors during stop. Check them and either stop
immediately (in default mode) or continue and report error in the
return value (in force mode). Add option -f to stop command to enable
force mode.

bird-tools/netlab/:
+ cf-babel-base/data uploaded.

bird-tools/netlab/cf-babel-base/bird_m*:
+ babel protocols now have option 'rxcost' with specific value
+ horizontal links - 100,
+ vertical links - 110 (m1-m2), 120 (m7-m8),
+ net1 - 150 (m3), 152 (m4), 154 (m5), 156 (m6).

With newer kernel/iproute, the output of 'ip route' on ECMP routes is
slightly different. The patch filter-out that as a temporary workaround.

bird-tools/netlab/:
+ cf-ospf-authentication/data updated,
+ cf-ospf-base/data updated,
+ cf-ospf-bfd/data updated,
+ cf-ospf-custom/data updated,
+ cf-ospf-default/data updated,
+ cf-ospf-nbma/data updated,
+ cf-ospf-priority/data updated,
+ cf-ospf-ptmp/data updated.

bird-tools/netlab/:
+ cf-ospf-authentication updated,
+ cf-ospf-base updated,
+ cf-ospf-bfd updated,
+ cf-ospf-custom updated,
+ cf-ospf-default updated,
+ cf-ospf-nbma updated,
+ cf-ospf-priority updated,
+ cf-ospf-ptmp updated.

bird-tools/netlab/:
+ cf-babel-base/data/* added,
+ cf-rip-base/data/* added.

Each test case contains:
+ Bird config files,
+ Netlab config file,
+ Python test suite.

VRF tables contain both regular and local routes, with older
kernels there are these routes:

  ff00::/8 dev eth0 metric 256 pref medium

while with newer ones there are:

  multicast ff00::/8 dev eth0 proto kernel metric 256 pref medium

Extend regex to ignore also routes from newer kernels.

bird-tools/netlab:
+ cf-ebgp-loop/data added
+ cf-ebgp-star/data added
+ cf-ibgp-flat/data added
+ cf-ibgp-loop/data added

+ test case 'cf-ebgp-loop' added
+ test case 'cf-ebgp-star' added
+ test case 'cf-ibgp-flat' added
+ test case 'cf-ibgp-loop' added

Each test case contains:
+ bird config files
+ netlab config file
+ Python test suite

As they are now stable.

As they are now stable.

Also fix pattern in cf-ospf-custom and some missing imports

+ function 'test_logs' allows some types of messages (DBG, TRACE, INFO),
  but the rest of the log messages should rase exceptions

+ function 'check_expected_logs_timeout' periodically calls function
  'check_expected_logs' which watches log messages every second

- functions 'modify_command', 'save_stdout', 'get_attributes_from_output',
  'check_running_protocols' and 'check_protocol_state' were removed