Age | Commit message (Collapse) | Author |
|
This reverts commit 4729a33cdf054cce6780aa4817f9eaccb490ee3f.
Need to put back the check, since it won't work asis
|
|
Bug 5383
Kernel modified to not default to vlan challenged, therefore
it is now okay to make vlan on empty bond
|
|
boundry
|
|
|
|
|
|
(cherry picked from commit 11baf27c6caf1a0edf2b993fb7066227d02eab7b)
|
|
(cherry picked from commit edf507d5be8061475d00c0cb3278f7cdd8ae844a)
|
|
to push users towards putting scripts in /config/scripts
|
|
deletion
|
|
community submitted patch see bug 5491
|
|
If commit cleans up temp file, no sudo is necessary.
|
|
Bug 6778
Both the nodes for speed and duplex both update the device.
But only the change only needs to be done once.
Changing the values twice causes link bounce and takes some
time to resolve.
|
|
|
|
Bug 6116
Do setup at same priority as VIF (since it is really macvlan)
Do address after peth mac address is set.
(cherry picked from commit 0f3aa38c15bb159d331dcfd91076ac37bd4c524e)
|
|
Bug 6642
Need slaves to be attached to bonding device before setting
address (for DHCP).
(cherry picked from commit 77297af866be32c058363f9cb3d104db9e00bb36)
|
|
is manually restared with "restart vrrp" command.
- Move all commit checks into script on end node.
- Add separate function for conntrack-sync settings.
- Do per instance validation at the right priority level.
- Verify that the changes file is removed on final transaction.
|
|
Bug 6489
Bonding device is a pseudo-device without hw-id, so deleting mac
does nothing.
|
|
DHCPv6 is not restricted to ethernet only anymore.
|
|
Bug 5864
The STP port priority field is limited to a smaller range to allow
more ports on the Linux bridge. Linux uses the same mapping as IOS
so that in the STP port field there are 10 bits for port id and
6 bits for port priority.
|
|
No longer need check-vip option since the type checking is done
in template
|
|
Use CLI (ipv4,ipv4net) and new address validator to check addresses for VRRP.
|
|
|
|
Bug 5673
Enfore restriction on values per RFC 3678
|
|
Speeds up boot with lots of tunnels
|
|
Avoid the overhead of perl compilation of vyatta-interfaces for each
address added. Handle the case of deleting address on deleted device
cleanly.
Clean up dhcpv6 script so it can be used on interfaces other
than ethernet.
|
|
|
|
Need to leave gre0 around for case of multiple tunnels
|
|
Bug 4608
|
|
This is necessary to get routing protocols and proper cleanup
|
|
Put in check for brN in bridge-group bridge.
And make syntax error for bond-group better
|
|
Move check for address and state of slave to commit which
allows for better error message.
|
|
Consolidate bridge port management into a single perl script.
Done to fix the following bug.
Bug 4745
Don't allow interface with address to be added to bridge.
|
|
Need to have same checks as ethernet for valid address on commit.
|
|
Comment was providing no useful information not already present in the
template.
|
|
Bug 5698
When interface transistions back to online, the driver
may have reacquired IRQ's and the irq affinity script will
need to be rerun.
|
|
Replace old script with new cleaner script that handles both
IRQ affinity and Receive Packet Steering. Instead of two scripts
(one for mask and one for auto), do it all with one script.
Receive Packet Steering is supported in two ways.
If 'auto' is used, then both threads on HT system will be
used for receive processing.
If explicit mask is given, then two masks can be used to set
both IRQ cpus and RPS cpus.
|
|
Bug 4695
If ethernet hardware does not allow VLAN's than produce error.
Ignore delete if device doesn't exist.
|
|
Bug 4994
Use ethtool to check for supported speed and duplex values on link.
|
|
The syntax of parent MTU was wrong. The underlying driver already
checks for MTU so the commit check is not needed.
|
|
Bug 6048
Preconditions:
310 bridge
315 bonding
320 bonding vif
325 bonding vif bridge group
|
|
Bug 3531
|
|
1. Move scripts to vyatta-cfg-system
2. Use syntax to check for legal values
3. Don't fail on UP
|
|
|
|
Setting bonding parameter requires setting interface down.
Avoid flapping interface unnecessarily on boot, by defering
setting link up until end of boot.
|
|
Turns out link must be down to change hash policy
|
|
|
|
|
|
(cherry picked from commit ce148b04d59fad50a26c394e4a59e252ea565e96)
|
|
Bug 5901
(cherry picked from commit 47014cdc10a429725d1620fc5326fd9995469c09)
|
|
The input interface needs to be in QoS package since the delete
action needs to refer to scripts in QoS package.
|