Internet DRAFT - draft-yizhou-trill-active-active-connection-prob
draft-yizhou-trill-active-active-connection-prob
TRILL Working Group Yizhou Li
INTERNET-DRAFT Donald Eastlake
Intended Status: Informational Weiguo Hao
Huawei Technologies
Radia Perlman
Intel Labs
Jon Hudson
Brocade
Hongjun Zhai
ZTE
Expires: August 18, 2014 February 14, 2014
Problem Statement and Goals for Active-Active TRILL Edge
draft-yizhou-trill-active-active-connection-prob-02
Abstract
The IETF TRILL (Transparent Interconnection of Lots of Links)
protocol provides support for flow level multi-pathing with rapid
failover for both unicast and multi-destination traffic in networks
with arbitrary topology between TRILL switches. Active-active at the
TRILL edge is the extension of these characteristics to end stations
that are multiply connected to a TRILL campus. This informational
document discusses the high level problems and goals when providing
active-active connection at the TRILL edge.
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as
Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/1id-abstracts.html
The list of Internet-Draft Shadow Directories can be accessed at
Yizhou, et al [Page 1]
INTERNET DRAFT Problems of Active-Active connection July 2013
http://www.ietf.org/shadow.html
Copyright and License Notice
Copyright (c) 2014 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Target Scenario . . . . . . . . . . . . . . . . . . . . . . . 3
3. Problems in Active-Active at the TRILL Edge . . . . . . . . . . 6
3.1 Frame Duplications . . . . . . . . . . . . . . . . . . . . . 6
3.2 Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Address Flip-Flop . . . . . . . . . . . . . . . . . . . . . 6
3.3 Unsynchronized Information Among Member RBridges . . . . . . 7
4 High Level Requirements and Goals for Solutions . . . . . . . . 7
5 Security Considerations . . . . . . . . . . . . . . . . . . . . 8
6 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 8
7 References . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7.1 Normative References . . . . . . . . . . . . . . . . . . . 8
7.2 Informative References . . . . . . . . . . . . . . . . . . 9
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 9
Yizhou, et al [Page 2]
INTERNET DRAFT Problems of Active-Active connection July 2013
1 Introduction
The IETF TRILL (Transparent Interconnection of Lots of Links)
[RFC6325] protocol provides loop free and per hop based multipath
data forwarding with minimum configuration. TRILL uses [IS-IS]
[RFC6165] [RFC6326bis] as its control plane routing protocol and
defines a TRILL specific header for user data. In a TRILL campus,
communications between TRILL switches can
(1) use multiple parallel links and/or paths,
(2) load spread over different links and/or paths at a fine grained
flow level through equal cost multipathing of unicast traffic and
multiple distribution trees for multi-destination traffic, and
(3) rapidly re-configure to accommodate link or node failures or
additions.
"Active-active" is the extension, to the extent practical, of similar
load spreading and robustness to the connections between end stations
and the TRILL campus. Such end stations may have multiple ports and
will be connected, directly or via bridges, to multiple edge TRILL
switches. It must be possible, except in some failure conditions, to
load spread end station traffic at the flow level across links to
such multiple edge TRILL switches and rapidly re-configure to
accommodate topology changes.
1.1 Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119 [RFC2119].
The acronyms and terminology in [RFC6325] is used herein with the
following additions:
CE - customer equipment. Could be a bridge or end station or a
hypervisor.
Edge group - a group of edge RBridges to which at least one CE is
multiply attached. One RBridge can be in more than one edge group.
TRILL switch - an alternative term for an RBridge.
2. Target Scenario
The TRILL appointed forwarder [RFC6325] [RFC6327bis] [RFC6439]
Yizhou, et al [Page 3]
INTERNET DRAFT Problems of Active-Active connection July 2013
mechanism provides per VLAN active-standby traffic spreading and loop
avoidance at the same time. One and only one appointed RBridge can
ingress/egress native frames into/from TRILL campus for a given VLAN
among all edge RBridges connecting a legacy network to TRILL campus.
This is true whether the legacy network is a simple point-to-point
link or a complex bridged LAN or anything in between. By carefully
selecting different RBridge as appointed forwarder for different set
of VLANs, load spreading over different edge RBidges across different
VLANs can be achieved.
This section presents a typical scenario of active-active connections
to TRILL campus via multiple edge RBridges where the current TRILL
appointed forwarder mechanism is not applicable.
The appointed forwarder mechanism [RFC6439] requires each of the edge
RBridges to exchange TRILL IS-IS Hello packets from their access
ports. As figure 1 shows, when multiple access links of multiple
edge RBridges are bundled as an MC-LAG (Multi-Chassis Link
Aggregation Group), Hello messages sent by RB1 via access port to CE1
will not be forwarded to RB2 by CE1. RB2 (and other members of MC-
LAG1) will not see that Hello from RB1. Every member RBridge of MC-
LAG1 thinks of itself as appointed forwarder on MC-LAG1 link for all
VLANs and will ingress/egress frames for all VLANs. Hence the
appointed forwarder mechanism is not applicable in such an active-
active scenario.
----------------------
| |
| TRILL Campus |
| |
----------------------
| | |
----- | --------
| | |
+------+ +------+ +------+
| | | | | |
|(RB1) | |(RB2) | | (RBk)|
+------+ +------+ +------+
|..| |..| |..|
| +----+ | | | |
| +---|-----|--|----------+ |
| +-|---|-----+ +-----------+ |
MC- | | | +------------------+ | |
LAG1--->(| | |) (| | |) <---MC-LAG2
+-------+ . . . +-------+
| CE1 | | CEn |
| | | |
+-------+ +-------+
Yizhou, et al [Page 4]
INTERNET DRAFT Problems of Active-Active connection July 2013
Active-Active connection is useful when we want to achieve the
following goals.
- Flow rather than VLAN based load balancing is desired.
- More rapid failure recovery is desired. Current appointed forwarder
mechanism relies on the Hello timer expiration to detect the
unreachability of another edge RBridge connecting to the same local
Ethernet link. Then re-appointing the forwarder for specific VLANs
may be required. Such procedures takes time in the scale of seconds.
Active-Active connection usually has faster built-in mechanism for
member node and/or link failure detection. Faster detection of
failure would minimize the frame loss and recovery time.
MC-LAG is a proprietary facility whose implementation varies by
vendor. So, to be sure of MC-LAG operation across an edge group of
RBridges, those edge RBridges will almost always be from the same
vendor. In order to have common understanding of active-active
connection scenarios, the following assumptions are made:
For CE connecting to multiple edge RBs via active-active connection:
a) the CE will forward a packet from an endnode to exactly one up-
link
b) the CE will never forward packets it receives from one up-link to
another
c) the CE will attempt to send all packets for a given flow on the
same uplink
d) packets are accepted from any of the uplinks and passed down to
endnodes (if any exist)
e) the CE has some unknown rule for which packets get sent to which
uplinks (typically based on a simple hash function of Layer 2 through
4 header fields)
f) the CE cannot be assumed to give useful control information to the
up-link such as "this set of other RBridges CE is attached", or
"these are all the MAC addresses attached"
For an edge group to which a CE is multiply attached:
a) Any two RBs in the edge group are reachable from each other
b) Each RB in the edge group is configured with a name for each down-
link to an CE multiply attached to that group. The names will be
consistent across the edge group. For instance, if CE1 attaches to
RB1, RB2 to RBn, then each of RBs will have been configured, for the
port to CE1, that it is labeled "MC-LAG1"
c) The RBs in the edge group have existing mechanisms to exchange
states and information with each other, including the set of CEs they
are connecting to or name of MC-LAGs their down-links have joined
d) Each RB in the edge group can be configured with the set of
Yizhou, et al [Page 5]
INTERNET DRAFT Problems of Active-Active connection July 2013
acceptable VLANs (or fine-grained labels) for the ports to any CE.
The acceptable VLANs configured for those port should include all the
VLANs the CE has joined and be consistent for all the member RB.
e) When a RB fails, all the other RBs having formed any MC-LAG with
it know the information timely
f) When a down-link of a RB fails, all the other RBs having formed
any MC-LAG with that down-link know the information timely
3. Problems in Active-Active at the TRILL Edge
This section presents the problems that need to be addressed in
active-active connection scenarios. The topology in Figure 1 is used
in the following sub-sections as the example scenario for
illustration purposes.
3.1 Frame Duplications
When a remote RBridge sends a multi-destination TRILL Data packet in
VLAN x, all member RBridges of MC-LAG1 will receive the frame if any
local CE1 joins VLAN x. As each of them thinks it is the appointed
forwarder for all VLANs, without active-active changes they would all
forward the frame to CE1. The bad consequence is that CE1 receives
multiple copies of that multi-destination frame from the remote end
host.
It should be noted frame duplication is only a problem in multi-
destination frame forwarding. Unicast forwarding does not have this
issue.
3.2 Loop
As shown in Figure 1, CE1 may send a native multi-destination frame
to TRILL campus via a member of MC-LAG1 (say RB1). This frame will be
TRILL encapsulated and then forwarded through the campus to another
member (say RB2) of the same MC-LAG. In this case, without active-
active changes RB2 will decapsulate the frame and forward it. The
frame loops back to CE1.
3.2 Address Flip-Flop
Consider RB1 and RB2 using their own nickname as ingress nickname for
data into a TRILL campus. As shown by Figure 1, CE1 may send a data
frame with the same source VLAN/MAC address to any member of the edge
group MC-LAG1. If the egress RBridge receives TRILL data packets from
different ingress RBridges but with same source VLAN/MAC address, it
learns different address correspondence from the decapsulated data
frames. Address correspondence may keep flip-flopping among nicknames
of the member RBridges of the MC-LAG for the same MAC address in the
Yizhou, et al [Page 6]
INTERNET DRAFT Problems of Active-Active connection July 2013
same VLAN.
Most TRILL switches may behave badly under these circumstances and,
for example, interpret this as a severe network problem. It may also
cause the returning traffic to go through the different paths to
reach the destination resulting in persistent re-ordering of the
frames.
3.3 Unsynchronized Information Among Member RBridges
A local Rbridge, say RB1 in MC-LAG1, may have learned a VLAN/MAC and
nickname correspondence for a remote host h1 when h1 sends a packet
to CE1. The returning traffic from CE1 may go to any other member
RBridge of MC-LAG1, e.g., RB2. RB2 may not have h1's VLAN/MAC and
nickname correspondence stored. Therefore it has to do the flooding
for unknown unicast. Such flooding is unnecessary since the returning
traffic is almost always expected and RB1 had learned the address
correspondence.
Synchronization on the VLAN/MAC and nickname correspondence
information among member RBridges will reduce such unnecessary
flooding.
Unsynchronized multicast group information causes problems too. The
edge RBridge snoops the IGMP [RFC3376] join message from CE may not
be the one receiving the multicast traffic for the joined group
later. Therefore multicast traffic can be dropped incorrectly.
TRILL [RFC6325] designed its multi-destination traffic forwarding
with some specific mechanisms, e.g., Reverse Path Forwarding Check,
tree calculation, construction and selection, pruning, etc. Solutions
of active-active connection at edge RBridges should carefully examine
those features and make sure they work correctly.
4 High Level Requirements and Goals for Solutions
Problems identified in section 3 should be solved in any solution for
active-active connection to RBridges. The requirements are summarized
as follows,
a) Loop and frame duplication MUST be prevented
b) Learning of VLAN/MAC and nickname correspondence by a remote
RBridge MUST not flip-flop between the local multiply attached edge
RBridges
c) Member RBridges of an MC-LAG MUST be able to share relevant TRILL
specific information with each other
In addition, the following high level goals should be met also.
Yizhou, et al [Page 7]
INTERNET DRAFT Problems of Active-Active connection July 2013
Data plane:
1) all up-links of CE MUST be active. CE is free to choose any up-
link on which to send packets
2) packets for a flow should stay in order
3) the Reverse Path Forwarding Check MUST work properly as per
[RFC6325]
4) Single up-link failure on CE to an edge group MUST not cause
persistent packet delivery failure between TRILL campus and CE
Control plane:
1) no requirement for new information to be passed between edge
RBridges and CE
2) If there are any TRILL specific parameters required to be
exchanged between RBridges in an edge group, e.g., nicknames,
solution SHOULD specify the mechanism to perform such exchange.
Configuration, incremental deployment and others:
1) Solution should require minimal configuration
2) Solution should automatically detect misconfiguration of edge
RBridge group
3) Solution should support incremental deployment, i.e. not require
campus wide upgrading for all RBridges, only changes to the edge
group RBridges
4) Solution should be able to support at least 4 active-active up-
links on a multiply attached CE
5 Security Considerations
This draft does not introduce any extra security risks. For general
TRILL Security Considerations, see [RFC6325].
6 IANA Considerations
No IANA action is required. RFC Editor: please delete this section
before publication.
7 References
7.1 Normative References
[IS-IS] ISO/IEC 10589:2002, Second Edition, "Intermediate System to
Intermediate System Intra-Domain Routing Exchange Protocol
for use in Conjunction with the Protocol for Providing the
Connectionless-mode Network Service (ISO 8473)", 2002.
Yizhou, et al [Page 8]
INTERNET DRAFT Problems of Active-Active connection July 2013
[RFC6165] Banerjee, A. and D. Ward, "Extensions to IS-IS for Layer-2
Systems", RFC 6165, April 2011.
[RFC6325] Perlman, R., Eastlake 3rd, D., Dutt, D., Gai, S., and A.
Ghanwani, "Routing Bridges (RBridges): Base Protocol
Specification", RFC 6325, July 2011
[RFC6326bis] Eastlake, D., Banerjee, A., Dutt, D., Perlman, R., and
A. Ghanwani, "TRILL Use of IS-IS", draft-eastlake-isis-
rfc6326bis, work in progress.
[RFC6327bis] Eastlake 3rd, D., R. Perlman, A. Ghanwani, H. Yang, and
V. Manral, "TRILL: Adjacency", draft-ietf-trill-
rfc6327bis, work in progress.
[RFC6439] Perlman, R., Eastlake, D., Li, Y., Banerjee, A., and F. Hu,
"Routing Bridges (RBridges): Appointed Forwarders", RFC
6439, November 2011
7.2 Informative References
[RFC3376] Cain, B., Deering, S., Kouvelas, I., Fenner, B., and A.
Thyagarajan, "Internet Group Management Protocol, Version
3", RFC 3376, October 2002.
[TRILLPN] Zhai,H., et.al., "RBridge: Pseudonode Nickname", draft-hu-
trill-pseudonode-nickname, Work in progress, November
2011.
[8021AX] IEEE, "Link Aggregration", 802.1AX-2008, 2008.
[8021Q] IEEE, "Media Access Control (MAC) Bridges and Virtual Bridged
Local Area Networks", IEEE Std 802.1Q-2011, August, 2011
Authors' Addresses
Yizhou Li
Huawei Technologies
101 Software Avenue,
Nanjing 210012
China
Phone: +86-25-56625409
EMail: liyizhou@huawei.com
Donald Eastlake
Yizhou, et al [Page 9]
INTERNET DRAFT Problems of Active-Active connection July 2013
Huawei Technologies
155 Beaver Street
Milford, MA 01757 USA
Phone: +1-508-333-2270
Email: d3e3e3@gmail.com
Weiguo Hao
Huawei Technologies
101 Software Avenue,
Nanjing 210012
China
Phone: +86-25-56623144
EMail: haoweiguo@huawei.com
Radia Perlman
Intel Labs
2200 Mission College Blvd.
Santa Clara, CA 95054-1549
USA
Phone: +1-408-765-8080
Email: Radia@alum.mit.edu
Jon Hudson
Brocade
130 Holger Way
San Jose, CA 95134 USA
Phone: +1-408-333-4062
jon.hudson@gmail.com
Hongjun Zhai
ZTE
68 Zijinghua Road, Yuhuatai District
Nanjing, Jiangsu 210012
China
Phone: +86 25 52877345
Email: zhai.hongjun@zte.com.cn
Yizhou, et al [Page 10]
