SYSTEM AND METHOD OF A HARDWARE SHADOW FOR A NETWORK ELEMENT

• US 20150006953A1
• Filed: 05/05/2014
• Published: 01/01/2015
• Est. Priority Date: 06/28/2013
• Status: Active Grant

• Current Assignee: Arista Networks, Inc.
• Original Assignee: Arista Networks, Inc.
• Inventors: Holbrook, Hugh W., Sellappa, Sriram
• Granted Patent: US 9,594,612 B2
• Time in Patent Office: Days
• US Class Current: 714/13
• CPC Class Codes: G06F 11/0724 in a multiprocessor or a multi-core unit multiprocessors per se G06F15/80
• CPC Class Codes: G06F 11/1658 Data re-synchronization of a redundant component, or initial sync of replacement, additional or spare unit
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups
• CPC Class Codes: H04L 69/321 Interlayer communication protocols or service data unit [SDU] definitions; Interfaces between layers
• CPC Class Codes: H04L 69/40 for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection management of faults, events, alarms or notifications in data switching networks H04L41/06

Abstract

A method and apparatus of a device that uses a hardware shadow for a central processing unit failover is described. In an exemplary embodiment, a device receives a signal that the active central processing unit has failed, where the active central processing unit controls the processing functions of the network element and the network element includes a hardware forwarding engine and a hardware shadow. The hardware shadow additionally includes multiple shadow tables. For each shadow table, the device copies data from that shadow table to a corresponding table in the hardware forwarding engine, where the hardware shadow is a copy of the data stored in hardware tables for the hardware forwarding engine. In response to the copying, the device further switches control of the network element processing functions from the active central processing unit to a standby central processing unit.

Claims

1. A non-transitory machine-readable medium having executable instructions to cause one or more processing units to perform a method to switch control of processing functions of a network element from an active central processing unit to a standby central processing unit, the method comprising:

receiving a signal that the active central processing unit has failed, wherein the active central processing unit controls the processing functions of the network element and the network element includes a hardware forwarding engine;

for each shadow table in a plurality of shadow tables in a hardware shadow, copying data from that shadow table to a corresponding one of a plurality of hardware tables for the hardware forwarding engine, wherein a hardware shadow is a copy of the data stored in the plurality of hardware tables for the hardware forwarding engine; and

in response to the copying, switching control of processing functions of the network element from the active central processing unit to a standby central processing unit.

2. The non-transitory machine-readable medium of claim 1, wherein the controlling of the network element processing functions with the standby central processing unit uses a shadow software index, wherein the shadow software index provides access to data of the hardware shadow for the standby central processing unit.

3. The non-transitory machine-readable medium of claim 2, further comprising:

reconciling the shadow software index with the hardware shadow.

4. The non-transitory machine-readable medium of claim 2, wherein the shadow software index includes data that is different from the data included in an active software index and the active software index provides access for the active central processing unit to the data stored in the plurality of hardware tables for the hardware forwarding engine.

5. The non-transitory machine-readable medium of claim 1, wherein the hardware shadow includes at least one of the plurality of shadow tables that is a permutation of a corresponding hardware table in the plurality of hardware tables.

6. The non-transitory machine-readable medium of claim 1, wherein the copying includes,

inserting a safe rule into the corresponding hardware table, wherein the safe rule allows a processing function associated with the corresponding hardware table to be in a consistent state.

7. The non-transitory machine-readable medium of claim 6, wherein the safe rule is selected from the group consisting of dropping traffic for that processing function, forwarding the traffic to a different forwarding engine, and forwarding the traffic to a central processing unit.

8. The non-transitory machine-readable medium of claim 1, wherein the copying further comprises:

copying the data into the corresponding hardware table such that the safe rule is not overwritten until the last entry of the corresponding hardware table is stored.

9. The non-transitory machine-readable medium of claim 1, further comprising:

signaling that the standby central processing unit is a new active central processing unit and has assumed control of the processing functions of the network element.

10. The non-transitory machine-readable medium of claim 1, wherein in response to the copying, the hardware shadow is synchronized with the plurality of hardware tables.

11. The non-transitory machine-readable medium of claim 1, wherein the copying data from that shadow table to a corresponding one of a plurality of hardware tables is performed asynchronously by one or more co-processors.

12. A non-transitory machine-readable medium having executable instructions to cause one or more processing units to perform a method to update a hardware shadow of a network element, the method comprising:

receiving a hardware shadow update, wherein the hardware shadow is used in a process of failing over control of processing functions of the network from an active central processing unit to a standby processing unit, the hardware shadow includes configuration data for one of a plurality of hardware forwarding engines of the network element, and the hardware shadow update is an update to one of a plurality of shadow tables of the hardware shadow; and

applying the update to the one of the plurality of hardware shadow tables.

13. The non-transitory machine-readable medium of claim 12, further comprising:

updating a shadow software index for the hardware shadow, wherein the shadow software index provides access to the plurality of shadow tables for the standby central processing unit.

14. The non-transitory machine-readable medium of claim 13, wherein the shadow software index includes data that is different from the data included in an active software index and the active software index provides access for an active central processing unit to the data stored in a plurality of hardware tables for the hardware forwarding engine.

15. The non-transitory machine-readable medium of claim 14, wherein the active central processing unit controls the network element processing functions and the standby central processing unit takes over the control of the network element processing functions in an event that the active central processing unit has a failure.

16. The non-transitory machine-readable medium of claim 13, wherein the hardware shadow includes a copy of the data that is stored in a plurality of hardware tables for the hardware forwarding engine.

17. The non-transitory machine-readable medium of claim 13, wherein the hardware shadow includes at least one of the plurality of shadow tables that is a permutation of a corresponding hardware table in the plurality of hardware tables.

18. The non-transitory machine-readable medium of claim 13, wherein the hardware shadow update is selected from the group consisting of an addition of an entry in one of the plurality shadow tables, a deletion of an entry in one of the plurality of shadow tables, and a modification to data of an entry in one of the plurality shadow tables.

19. The non-transitory machine-readable medium of claim 13, wherein one of the plurality of shadow tables is selected from the groups consisting of a routing table, a media access control table, and an access control list.

20. A non-transitory machine-readable medium having executable instructions to cause one or more processing units to perform a method to switch control of processing functions of a network element from an active central processing unit to a standby central processing unit, the method comprising:

receiving a signal that the active central processing unit has failed, wherein the active central processing unit controls the processing functions of the network element, and the network element includes a hardware forwarding engine, a hardware shadow and a partial software index for the standby central processing unit;

for each hardware table in the plurality of hardware tables referenced in the partial software index, switching control to the standby central processing unit of a processing function of the network element that utilizes data stored in that hardware table; and

for each shadow table in a plurality of shadow tables in the hardware shadow,

copying data from that shadow table to a corresponding one of a plurality of hardware tables in the hardware forwarding engine, wherein the hardware shadow is a copy of the data stored in the plurality of hardware tables for the hardware forwarding engine and not referenced by the partial software index, and

in response to the copying, switching control to the standby central processing unit of a processing function of the network element that utilizes data associated with that shadow table.

21. A non-transitory machine-readable medium having executable instructions to cause one or more processing units to perform a method to update a hardware shadow or a partial software index of a network element, the method comprising:

receiving an update, wherein the update is an update for one of the hardware shadow and the partial software index, the hardware shadow is used to failover a control of processing functions of the network from an active central processing unit to a standby processing unit, the hardware shadow includes configuration data for one of a plurality of hardware forwarding engines of the network element, and the partial software index references data stored in a plurality of hardware tables and not stored in the hardware shadow;

determining if the update is for the hardware shadow or the partial software index; and

if the update is for the hardware shadow,

applying the update to the hardware shadow update for the one of the plurality of hardware shadow tables; and

if the update is for the partial software index,

applying the update to the partial software index.