| Diameter Maintenance and Extensions (DIME) | S. Donovan |
| Internet-Draft | Oracle |
| Intended status: Standards Track | August 31, 2015 |
| Expires: March 3, 2016 |
Diameter Agent Overload
draft-ietf-dime-agent-overload-02.txt
This specification documents an extension to the Diameter Overload Indication Conveyance (DOIC) base solution. The extension addresses the handling of occurrences of overload of a Diameter agent, or more generally, a Diameter peer.
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].
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on March 3, 2016.
Copyright (c) 2015 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.
This document defines the behavior of Diameter nodes when Diameter agents enter an overload condition and send an overload report requesting a reduction of traffic.
The base Diameter overload specification [I-D.ietf-dime-ovli] addresses the handling of overload when a Diameter endpoint (a Diameter Client or Diameter Server as defined in [RFC6733]) becomes overloaded.
In the base specification, the goal is to handle abatement of the overload occurrence as close to the source of the Diameter traffic as is feasible. When possible this is done at the originator of the traffic, generally referred to as a Diameter Client. A Diameter Agent might also handle the overload mitigation. For instance, a Diameter Agent might handle Diameter overload mitigation when it knows that a Diameter Client does not support the DOIC extension.
This document extends the base Diameter endpoint overload specification to address the case when Diameter Agents become overloaded. Just as is the case with other Diameter nodes -- Diameter Clients and Diameter Servers -- surges in Diameter traffic can cause a Diameter Agent to be asked to handle more Diameter traffic than it was configured to handle. For a more detailed discussion of what can cause the overload of Diameter nodes, refer to the Diameter Overload Requirements [RFC7068].
This document defines a new overload report type to communicate occurrences of agent overload. This report type works for the "Loss" overload mitigation algorithm defined in [I-D.ietf-dime-ovli] and is expected to work for other overload abatement algorithms defined in extensions to the DOIC solution.
The handling of endpoint overload and agent overload is very similar. The primary differences are the following:
Editor's Note: Open Issue - Does a peer report apply to the implicitly communicated application-id in the same way as host and realm reports do or does it apply to all applications handled by the peer? Do we need the ability for to support both cases?
Open Issue - To support the ability of an agent to select a different abatement algorithm than endpoints, we probably need to extend the OC-Supported-Features AVP to include an OC-Abatement-Algorithm AVP. This is currently shown to be in the OC-OLR AVP but needs to be moved as this information is needed prior to receiving the OC-OLR. It probably needs to be changed to OC-Peer-Abatement-Algorithm.
Editors note – These definitions need to be made consistent with the base Diameter overload specification defined in [I-D.ietf-dime-ovli].
This section outlines representative use cases for the peer report used to communicate agent overload.
There are two primary classes of use cases, those involving the overload of agents and those involving overload of Diameter endpoints (Diameter Clients and Diameter Servers).
The agent overload extension must support the following use cases.
This use case is illustrated in Figure 1. In this case, the client sends all traffic through the single agent. If there is a failure in the agent then the client is unable to send Diameter traffic toward the server.
+-+ +-+ +-+
|c|----|a|----|s|
+-+ +-+ +-+
Figure 1
A more likely case for the use of agents is illustrated in Figure 2. In this case, there are multiple servers behind the single agent. The client sends all traffic through the agent and the agent determines how to distribute the traffic to the servers based on local routing and load distribution policy.
+-+
--|s|
+-+ +-+ / +-+
|c|----|a|- ...
+-+ +-+ \ +-+
--|s|
+-+
Figure 2
In both of these cases, the occurrence of overload in the single agent must by handled by the client in a similar fashion as if the client were handling the overload of a directly connected server. When the agent becomes overloaded it will insert an overload report in answer messages flowing to the client. This overload report will contain a requested reduction in the amount of traffic sent to the agent. The client will apply overload abatement behavior as defined in the base Diameter overload specification [I-D.ietf-dime-ovli] or the extension draft that defines the indicated overload abatement algorithm. This will result in the throtting of the abated traffic that would have been sent to the agent, as there is no alternative route, with the appropriate indication given to the service request that resulted in the need for the Diameter transaction.
Figure 3 and Figure 4 illustrate a second, and more likely, type of deployment scenario involving agents. In both of these cases, the client has Diameter connections to two agents.
Figure 3 illustrates a client that has a primary connection to one of the agents (agent a1) and a secondary connection to the other agent (agent a2). In this scenario, under normal circumstances, the client will use the primary connection for all traffic. The secondary connection is used when there is a failure scenario of some sort.
+--+ +-+
--|a1|---|s|
+-+ / +--+\ /+-+
|c|- x
+-+ . +--+/ \+-+
..|a2|---|s|
+--+ +-+
Figure 3
The second case, in Figure 4, illustrates the case where the connections to the agents are both actively used. In this case, the client will have local distribution policy to determine the percentage of the traffic sent through each client.
+--+ +-+
--|a1|---|s|
+-+ / +--+\ /+-+
|c|- x
+-+ \ +--+/ \+-+
--|a2|---|s|
+--+ +-+
Figure 4
In the case where one of the agents in the above scenarios become overloaded, the client should reduce the amount of traffic sent to the overloaded agent by the amount requested. This traffic should instead be routed through the non-overloaded agent. For example, assume that the overloaded agent requests a reduction of 10 percent. The client should send 10 percent of the traffic that would have been routed to the overloaded agent through the non-overloaded agent.
When the client has an active and a standby connection to the two agents then an alternative strategy for responding to an overload report from an agent is to change to standby connection to active and route all traffic through the new active connection.
In the case where both agents are reporting overload, the client may need to start decreasing the total traffic sent to the agents. This would be done in a similar fashion as discussed in section 3.1. The amount of traffic depends on the combined reduction requested by the two agents.
There are also deployment scenarios where there can be multiple Diameter Agents between Diameter Clients and Diameter Servers. Examples of this type of deployment include when there are edge agents between Diameter networks. Another example of this type of deployment is when there are multiple sets of servers, each supporting a subset of the Diameter traffic.
Figure 5 illustrates one such network deployment case. Note that while this figure shows a maximum of two agents being involved in a Diameter transaction, it is possible that more than two agents could be in the path of a transaction.
+---+ +---+ +-+
--|a11|-----|a21|---|s|
+-+ / +---+ \ / +---+\ /+-+
|c|- x x
+-+ \ +---+ / \ +---+/ \+-+
--|a12|-----|a22|---|s|
+---+ +---+ +-+
Figure 5
Handling of overload of one or both of agents a11 or a12 in this case is equivalent to that discussed in section 2.2.
Overload of agents a21 and a22 must be handled by the previous hop agents. As such, agents a11 and a12 must handle the overload mitigation logic when receiving an agent overload report from agents a21 and a22.
Editor's note: Probably need to elaborate the reasoning behind the need for the agent overload report being handled by the previous hop agent.
The handling of peer overload reports is similar to that discussed in section 2.2. If the overload can be addressed using diversion then this approach should be taken.
If both of the agents have requested a reduction in traffic then the previous hop agent must start throttling the appropriate number of transactions. When throttling requests, an agent uses the same error responses as defined in the base DOIC specification [I-D.ietf-dime-ovli].
This section outlines use cases for the peer report feature involving Diameter Clients and Diameter Servers.
It is envisioned that abatement algorithms will be defined that will support the option for Diameter Endpoints to send peer reports. For instance, it is envisioned that one usage scenario for the rate algorithm, [I-D.ietf-dime-doic-rate-control], which is being worked on by the DIME working group as this is written, will involve abatement being done on a hop-by-hop basis.
This rate deployment scenario would involve Diameter Endpoints generating peer reports and selecting the rate algorithm for abatement of overload conditions.
It is possible that both an agent and a server in the path of a transaction are overloaded at the same time. When this occurs, Diameter entities will need to handle both overload reports. In this scenario the reacting node should first handle the throttling of the overloaded host or realm. Any messages that survive throttling due to host or realm reports should then go through abatement for the peer overload report.
This section defines the normative behavior associated with the Peer Report extension to the DOIC solution.
Editor's Note: Issue - how does an agent indicate the selected abatement algorithm? It cannot use the OC-Feature-Vector in the OC-Supported-Features AVP as that applies to host and realm report types. A new AVP in the OC-Supported-Features AVP has been added.
When sending a Diameter request a DOIC node that supports the Peer Report feature MUST include an OC-Supported-Features AVP with an OC-Feature-Vector AVP with the OLR_PEER_REPORT bit set.
Support for the peer report feature does not impact the logic for setting of other feature bits in the OC-Feature-Vector AVP.
When sending a request a DOIC node that supports the Peer Report feature MUST include an OC-SourceID AVP in the OC-Supported-Features AVP with its own DiameterID.
When receiving a request a DOIC node that supports the Peer Report feature MUST update transaction state with an indication of whether or not the peer from which the request was received supports the Peer Report feature.
The following are indications that the peer does not support the OLR_PEER_REPORT feature:
The peer supports the OLR_PEER_REPORT feature if the received request contains an OC-Supported-Features AVP with the OC-Feature-Vector with the OLR_PEER_REPORT feature bit set and with an OC-SourceID AVP with a Diameter ID that matches the DiameterID of the peer from which the request was received.
When receiving a request a DOIC node that supports the Peer Report feature MUST remove any received OC-SourceID AVP from the OC-Supported-Features AVP. This is done to prevent the OC-SourceID AVP from being included in a relayed message through a node that supports the Peer Report feature.
When sending an answer message, a reporting node that supports the OLR_PEER_REPORT feature MUST strip any SourceID AVP from the OC-Supported-Features AVP.
When sending an answer message, a reporting node that supports the OLR_PEER_REPORT feature MUST determine if the peer to which the answer is to be sent supports the OLR_PEER_REPORT feature.
If the peer supports the OLR_PEER_REPORT feature then the reporting node MUST indicate support for the feature in the Supported-Features AVP.
If the peer supports the OLR_PEER_REPORT feature then the reporting node MUST insert the OC-SourceID AVP in the OC-Supported-Features AVP in the answer message.
If the peer supports the OLR_PEER_REPORT feature then the reporting node MUST insert the OC-Peer-Algo AVP in the OC-Supported-Features AVP. The OC-Peer-Algo AVP MUST indicate the overload abatement algorithm that the reporting node wants reacting nodes to use should the reporting node send a peer overload report as a result of becoming overloaded.
This section defines the behavior for the handling of overload reports of type peer.
This section describes the Overload Control State (OCS) that might be maintained by both the peer report reporting node and the peer report reacting node.
A DOIC Node that supports the Peer Report feature SHOULD maintain Reporting Node Peer Report OCS. This is used to record overload events and build overload reports at the reporting node.
If different abatement specific contents are sent to each peer then the reporting node MUST maintain a separate peer node peer report OCS entry per peer to which a peer overload report is sent.
The Reporting Node Peer Report OCS entry MAY include the following information (the actual information stored is an implementation decision):
A DOIC node that supports the Peer Report feature SHOULD maintain Reacting Node Peer Report OCS for each peer with which it communicates. This is used to record overload reports received from peer nodes.
A Reacting Node Peer Report OCS entry is identified by the DiameterID of the peer as communicated during the RFC6733 defined Capability Exchange procedure
The Reacting Node Peer Report OCS entry MAY include the following information (the actual information stored is an implementation decision):
A reporting node SHOULD create a new Reporting Node Peer Report OCS entry Section 5.2.1.1 in an overload condition and sending a peer overload report to a peer for the first time.
All rules for managing the reporting node OCS entries defined in [DOIC] apply to the peer report.
When a reacting node receives an OC-OLR AVP with an a report type of peer it MUST determine if the report was generated by the Diameter peer from which the report was received.
If the DiameterID in the SourceID contained in the OLR matches the DiameterID of the peer from which the request was received then the report was received from a Diameter peer.
If a reacting node receives an OC-OLR AVP of type peer and the OC-SourceID does not match the ID of the Diameter peer from which the request was received then the reacting node MUST strip the OC-OLR AVP from the message and not use it to update reacting node peer report OCS entries.
If the Peer Report OLR was received from a Diameter peer then the reacting node MUST determine if it is for an existing or new overload condition.
The OLR is for an existing overload condition if the reacting node has an OCS that matches the received OLR.
For a peer report-type this means the DiameterID received in the SourceID AVP matches the DiameterID of an existing peer report OLR.
If the OLR is for an existing overload condition then it MUST determine if the OLR is a retransmission or an update to the existing OLR.
If the sequence number for the received OLR is greater than the sequence number stored in the matching OCS entry then the reacting node MUST update the matching OCS entry.
If the sequence number for the received OLR is less than or equal to the sequence number in the matching OCS entry then the reacting node MUST silently ignore the received OLR. The matching OCS MUST NOT be updated in this case.
If the received OLR is for a new overload condition then the reacting node MUST generate a new OCS entry for the overload condition.
For a peer report this means it creates an OCS entry with an DiameterID from the SourceID AVP in the received OC-OLR AVP.
If the received OLR contains a validity duration of zero ("0") then the reacting node MUST update the OCS entry as being expired.
The reacting node does not delete an OCS when receiving an answer message that does not contain an OC-OLR AVP (i.e. absence of OLR means "no change").
The reacting node sets the abatement algorithm based on the OC-Peer-Algo AVP in the received OC-Supported-Features AVP.
When there is an existing reporting node peer report OCS entry, the reporting node MUST include an OC-OLR AVP with a report type of peer using the contents of the reporting node peer report OCS entry in all answer messages sent by the reporting node to peers that support the peer report feature.
The reporting node MUST include its DiameterID in the OC-SourceID AVP in the OC-OLR AVP. This is used by DOIC nodes that support the peer report feature to determine if the report was received from a Diameter peer.
The reporting agent must follow all other overload reporting node behaviors outlined in the DOIC specification.
A reacting node supporting this extension MUST support the receipt of multiple overload reports in a single message. The message might include a host overload report, a realm overload report and a peer overload report.
When a reacting node sends a request it MUST determine if that request matches an active OCS.
If the request matches and active OCS then the reacting node MUST apply abatement treatment on the request. The abatement treatment applied depends on the abatement algorithm stored in the OCS.
For peer overload reports, the preferred abatement treatment is diversion. As such, the reacting node SHOULD attempt to divert requests identified as needing abatement to other peers.
If a host-routed request, as defined in [I-D.ietf-dime-ovli], is selected for abatement and the request must be routed to the DOIC node that generated the peer overload report -- meaning that the request is a host-routed request as defined in the DOIC specification -- then the reacting node MUST throttle the request.
If there is not sufficient capacity to divert abated traffic then the reacting node MUST throttle the necessary requests to fit within the available capacity of the peers able to handle the requests.
If the abatement treatment results in throttling of the request and if the reacting node is an agent then the agent MUST send an appropriate error as defined in [I-D.ietf-dime-ovli].
In the case that the OCS entry validity duration expires or has a validity duration of zero ("0"), meaning that it the reporting node has explicitly signaled the end of the overload condition then abatement associated with the overload abatement MUST be ended in a controlled fashion.
This extension adds a new feature to the OC-Feature-Vector AVP. This feature indication shows support for handling of peer overload reports. Peer overload reports are used by agents to indicate the need for overload abatement handling by the agents peer.
A supporting node must also include the OC-SourceID AVP in the OC-Supported-Features capability AVP.
This AVP contains the Diameter Identity of the node that supports the OLR_PEER_REPORT feature. This AVP is used to determine if support for the peer overload report is in an adjacent node. The value of this AVP should be the same Diameter identity used as part of the CER/CEA base Diameter capabilities exchange.
OC-Supported-Features ::= < AVP Header: TBD1 >
[ OC-Feature-Vector ]
[ OC-SourceID ]
[ OC-Peer-Algo]
* [ AVP ]
The peer report feature defines a new feature bit is added for the OC-Feature-Vector AVP.
The OC-Peer-Algo AVP (AVP code TBD6) is of type Unsigned64 and contains a 64 bit flags field of announced capabilities of a DOIC node. The value of zero (0) is reserved.
Feature bits defined for the OC-Feature-Vector AVP and associated with overload abatement algorithms are reused in for this AVP.
This extension makes no changes to the SequenceNumber or ValidityDuration AVPs in the OC-OLR AVP. These AVPs are also be used in peer overload reports.
The peer report feature extends the base Diameter overload specification by defining a new overload report type of “peer”. See section [7.6] in [I-D.ietf-dime-ovli] for a description of the OC-Report-Type AVP.
The overload report must also include the Diameter identity of the agent that generated the report. This is necessary to handle the case where there is a non supporting agent between the reporting node and the reacting node. Without the indication of the agent that generated the overload request, the reacting node could erroneously assume that the report applied to the non supporting node. This could, in turn, result in unnecessary traffic being either redistributed or throttled.
The OC-SourceID AVP is used in the OC-OLR AVP to carry this DiameterID.
OC-OLR ::= < AVP Header: TBD2 >
< OC-Sequence-Number >
< OC-Report-Type >
[ OC-Reduction-Percentage ]
[ OC-Validity-Duration ]
[ OC-Source-ID ]
* [ AVP ]
The following new report type is defined for the OC-Report-Type AVP.
The SourceID AVP (AVP code TBD) is of type DiameterIdentity and is inserted by the DOIC node that either indicates support for this feature (in the OC-Supported-Features AVP) or that generates an OC-OLR AVP with a report type of peer.
It contains the Diameter Identity of the inserting node. This is used by other DOIC nodes to determine if the a peer indicated support this feature or inserted the peer report.
+---------+
|AVP flag |
|rules |
+----+----+
AVP Section | |MUST|
Attribute Name Code Defined Value Type |MUST| NOT|
+--------------------------------------------------------+----+----+
|OC-SourceID TBD1 x.x Unsigned64 | | V |
|OC-Peer-Algo TBD1 x.x Unsigned64 | | V |
+--------------------------------------------------------+----+----+
Editors note: This section will be completed once the base overload document has finished the definition of extension IANA requirements.
Agent overload is an extension to the based Diameter overload mechanism. As such, all of the security considerations outlined in [I-D.ietf-dime-ovli] apply to the agent overload scenarios.
It is possible that the malicious insertion of an agent overload report could have a bigger impact on a Diameter network as agents can be concentration points in a Diameter network. Where an end-point report would impact the traffic sent to a single Diameter server, for example, a peer report could throttle all traffic to the Diameter network.
This impact is amplified in an agent that sits at the edge of a Diameter network that serves as the entry point from all other Diameter networks.
Adam Roach and Eric McMurry for the work done in defining a comprehensive Diameter overload solution in draft-roach-dime-overload-ctrl-03.txt.
Ben Campbell for his insights and review of early versions of this document.
| [I-D.ietf-dime-doic-rate-control] | Donovan, S. and E. Noel, "Diameter Overload Rate Control", Internet-Draft draft-ietf-dime-doic-rate-control-01, March 2015. |
| [I-D.ietf-dime-ovli] | Korhonen, J., Donovan, S., Campbell, B. and L. Morand, "Diameter Overload Indication Conveyance", Internet-Draft draft-ietf-dime-ovli-08, February 2015. |
| [RFC2119] | Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997. |
| [RFC5226] | Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 5226, DOI 10.17487/RFC5226, May 2008. |
| [RFC6733] | Fajardo, V., Arkko, J., Loughney, J. and G. Zorn, "Diameter Base Protocol", RFC 6733, DOI 10.17487/RFC6733, October 2012. |
| [RFC7068] | McMurry, E. and B. Campbell, "Diameter Overload Control Requirements", RFC 7068, DOI 10.17487/RFC7068, November 2013. |