From Fedora Project Wiki
 
(19 intermediate revisions by the same user not shown)
Line 5: Line 5:
=== Kernel ===
=== Kernel ===


* [[http://bugzilla.redhat.com/show_bug.cgi?id=891214 Ethernet: noarp causes dst MAC to be the same as src MAC instead of broadcast]]
* [http://bugzilla.redhat.com/show_bug.cgi?id=891214 Ethernet: noarp causes dst MAC to be the same as src MAC instead of broadcast]
* Bridge: Adding first interface after setting MAC address changes the MAC address
* [http://bugzilla.redhat.com/show_bug.cgi?id=886617 Bridge/bond: Empty bond interface cannot be bridged unless you work around by manually setting the address]
* Bridge/bond: [https://bugzilla.redhat.com/show_bug.cgi?id=886617 Empty bond interface cannot be bridged unless you work around by manually setting the address]
* [http://bugzilla.redhat.com/show_bug.cgi?id=905868 Kernel removes ::1 but not 127.0.0.1 when setting lo down]
* Dualstack: Kernel removes ::1 but not 127.0.0.1 when setting lo down
* [http://bugzilla.redhat.com/show_bug.cgi?id=891234 IPv4: Kernel can't store DHCP lifetime and cannot mark IPv4 addresses dynamic (in the DHCP sense)] (due in 3.9)
* IPv4: Kernel can't store DHCP lifetime and cannot mark IPv4 addresses dynamic (in the DHCP sense)
* [http://bugzilla.redhat.com/show_bug.cgi?id=886116 IPv6: IPv6 networking breaks when loopback is set down and up]
* IPv6: [https://bugzilla.redhat.com/show_bug.cgi?id=886116 IPv6 networking breaks when loopback is set down and up]
* IPv6: Kernel ignores RA router preference
* IPv6: Kernel ignores RA router preference
* IPv6: Tempaddr cycling breaks TCP connections prematurely
* IPv6: Tempaddr cycling breaks TCP connections prematurely
* IPv6: Kernel doesn't cache RDNSS and DNSSL
* [http://bugzilla.redhat.com/show_bug.cgi?id=906505 IPv6: API for router discovery]
* IPv6: Kernel doesn't provide default route information when defrtr is disabled
* [http://bugzilla.redhat.com/show_bug.cgi?id=905880 IPv6: Kernel doesn't cache RDNSS and DNSSL]
* IPv6: Kernel doesn't provide prefix route information when pinfo is disabled
* [http://bugzilla.redhat.com/show_bug.cgi?id=891245 IPv6: Kernel doesn't provide managable IPv6 autoconfiguration]
* Netlink: Kernel doesn't solve race condition between rtnl commands and events
* ?Netlink: Kernel doesn't solve race condition between rtnl commands and events
* Netlink: Kernel doesn't send some link change notifications (e.g. mtu) for devices that are down
* <del>Netlink: Kernel doesn't send some link change notifications (e.g. mtu) for devices that are down</del>
* master/slave: Kernel enslaves a L3-configured interface even if it can't work that way (the same in the opposite order, reported by djasa)
* bridge: re-enslaving a slave to another bridge may cause two bridges to have duplicate MACs
* [http://bugzilla.redhat.com/show_bug.cgi?id=910348 bridge: an empty bridge gets IFF_LOWER_UP when set up, which constitutes a race condition for carrier detection when first slave is added]
* [http://bugzilla.redhat.com/show_bug.cgi?id=910810 bridge: an empthy bridge doesn't get IFF_MASTER]


=== libnl ===
=== libnl ===
Line 39: Line 42:


[[Networking/NameResolution/ADDRCONFIG|More information about AI_ADDRCONFIG problems]]
[[Networking/NameResolution/ADDRCONFIG|More information about AI_ADDRCONFIG problems]]
[http://sourceware.org/bugzilla/buglist.cgi?quicksearch=getaddrinfo Upstream bug reports for getaddrinfo]


=== dhclient ===
=== dhclient ===


* Fixed: [http://bugzilla.redhat.com/show_bug.cgi?id=662254 <code>dhclient</code> fails with large values of dhcp-renawal-time, notably the infinity (0xffffffff) value]
* <del>[http://bugzilla.redhat.com/show_bug.cgi?id=662254 <code>dhclient</code> fails with large values of dhcp-renawal-time, notably the infinity (0xffffffff) value]</del>
* People say that dhclient exchange is slow, especially on wireless (I can't reproduce)
* People say that dhclient exchange is slow, especially on wireless (I can't reproduce)


Line 49: Line 54:
* [http://bugzilla.redhat.com/show_bug.cgi?id=719178 Avahi's ''nss-mdns'' returns link-local IPv6 addresses ''without'' scope_id and applications fail to connect]
* [http://bugzilla.redhat.com/show_bug.cgi?id=719178 Avahi's ''nss-mdns'' returns link-local IPv6 addresses ''without'' scope_id and applications fail to connect]
* [http://bugzilla.redhat.com/show_bug.cgi?id=821127 Avahi's IPv6 support is disabled by default (this has been resolved in the past)]
* [http://bugzilla.redhat.com/show_bug.cgi?id=821127 Avahi's IPv6 support is disabled by default (this has been resolved in the past)]
See [[Features/ZeroconfNetworking]].


=== NetworkManager ===
=== NetworkManager ===
Line 56: Line 63:
=== Testing tools ===
=== Testing tools ===


* ping doesn't support IPv6 (you have to use ping6)
* [http://bugzilla.redhat.com/show_bug.cgi?id=617934 ping doesn't support IPv6 (you have to use ping6)]
* netstat and ss won't mark non-v6only sockets
* netstat and ss won't mark non-v6only sockets


Line 63: Line 70:
* Assume that <code>getaddrinfo()</code> returns only one address and/or that such the first returned address works. See [[Networking/NameResolution]].
* Assume that <code>getaddrinfo()</code> returns only one address and/or that such the first returned address works. See [[Networking/NameResolution]].


=== Network related standards ===
=== Networking related standards ===


* RFC 6106 (RA DNS): [http://tools.ietf.org/html/draft-gont-6man-slaac-dns-config-issues-00 Relies on a false assuption that every (even multicast) IP packet is delivered to its destination.]
* RFC 6106 (RA DNS): [http://tools.ietf.org/html/draft-gont-6man-slaac-dns-config-issues-00 Relies on a false assuption that every (even multicast) IP packet is delivered to its destination.]
Line 69: Line 76:
* RFC 3493 (socket interfce): Breaks both IPv4 and IPv6 localhost and link-local networking depending on available global addresses.
* RFC 3493 (socket interfce): Breaks both IPv4 and IPv6 localhost and link-local networking depending on available global addresses.
* POSIX.1-2008: [https://bugzilla.redhat.com/show_bug.cgi?id=721350#c8 Defines the AI_ADDRCONFIG in a bad way]
* POSIX.1-2008: [https://bugzilla.redhat.com/show_bug.cgi?id=721350#c8 Defines the AI_ADDRCONFIG in a bad way]
* RFC 4862 (SLAAC) says prefix with AdvAutonomous off should be ignored (routing would fail with DHCP then)
== Contacts ==
See [[Networking#Contacts]].

Latest revision as of 14:20, 15 March 2013

Known networking bugs

Only platform bugs that can influence various application and/or the overall behavior of the system belong here.

Kernel

libnl

  • rtnl_link_bond_add() succeeds when device already exists
  • rtnl_link_bridge_add() and rtnl_link_vlan_add missing

iproute

  • ip link show won't distinguish bridges and bonds
  • ip link show won't show vlan id
  • ip -6 route shows cached routes (ip -4 route doesn't)

glibc

More information about AI_ADDRCONFIG problems

Upstream bug reports for getaddrinfo

dhclient

Avahi

See Features/ZeroconfNetworking.

NetworkManager

Testing tools

Various applications

  • Assume that getaddrinfo() returns only one address and/or that such the first returned address works. See Networking/NameResolution.

Networking related standards

Contacts

See Networking#Contacts.