Non-invasive latency monitoring in a store-and-forward replication system

• US 20060168120A1
• Filed: 02/17/2006
• Published: 07/27/2006
• Est. Priority Date: 11/27/2001
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Parham, Jeffrey B.
• Granted Patent: US 7,519,736 B2
• Time in Patent Office: Days
• US Class Current: 709/218
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/202 where processing functionality is redundant redundant communication control functionality G06F11/2005, redundant storage control functionality G06F11/2089
• 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: G06F 16/273 Asynchronous replication or reconciliation
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A method for monitoring replica servers in a networked computer system is provided, in which each server in the system has a replica partner vector table that includes state information about other servers in the system. The replica partner vector table includes data fields for storing an update sequence number (USN) and timestamp information that identifies the time of the last update and/or the time last successful replication attempt for each replica server in the system. After each successful replication, the server updates the entries in the replica partner vector to reflect the updated USN and timestamp information. The replica monitoring method evaluates the USN and timestamp entries in the replica partner vector table to determine if any servers in the system are latent. If the monitoring method detects that a server in the system is latent, an alert is generated whereby users and/or a network administrator are informed of the problem.

Claims

1. A method for monitoring replication latency in a computer system comprising a plurality of servers connected by a plurality of data links, and wherein the servers periodically replicate object updates with one another and maintain an update sequence number that is increased upon update to the server'"'"'s replica of a database, the method comprising:

first maintaining, by each of the plurality of servers, a timestamp that is posted upon each update to the server'"'"'s database;

second maintaining, by two or more of the plurality of servers, a replica partner vector table that includes for each other server from which the server replicates, the update sequence number of such other server at a time of a most recent replication from such other server or the timestamp of the last successful replication attempt with such other server;

transmitting a copy of the replica partner vector table of a local server to a remote server;

comparing, by the remote server, the update sequence numbers and timestamps in the replica partner vector table received from the local server to the update sequence numbers and timestamps in the replica partner vector table of the remote server;

updating, by the remote server, the replica partner vector table received from the local server;

transmitting, by the remote server, a copy of the updated replica partner vector table and object updates to the local server;

improving, by the local server, upon receiving the updated replica partner vector from the remote server, update sequence numbers or timestamps in the partner vector table of the local server; and

calculating, by the local server, a difference between the timestamp for each server in the replica partner vector and a current time.

2. The method of claim 1, further comprising:

transmitting a copy of the remote server'"'"'s replica partner vector table to the local server;

comparing the remote server'"'"'s replica partner vector table with the local server'"'"'s replica partner vector table; and

updating the replica partner vector table received from the remote server; and

transmitting a copy of the updated replica partner vector table and object updates to the remote server.

3. A computer-readable medium having computer-executable components thereon for monitoring replication latency in a computer system, the computer system comprising a plurality of servers connected by a plurality of data links, and wherein the servers periodically replica object updates with one another and maintain an update sequence number that is increased upon update to the server'"'"'s replica of a database, the components performing the steps of:

first maintaining, by each of the plurality of servers, a timestamp that is posted upon each update to the server'"'"'s database;

second maintaining, by two or more of the plurality of servers, a replica partner vector table that includes for each other server from which the server replicates, the update sequence number of such other server at a time of a most recent replication from such other server or the timestamp of the last successful replication attempt with such other server;

transmitting a copy of the replica partner vector table of a local server to a remote server;

comparing, by remote server, the update sequence numbers and timestamps in the replica partner vector table received from the local server to the update sequence numbers or timestamps in the replica partner vector table of the remote server;

updating, by the remote server, the replica partner vector table received from the local server;

transmitting, by the remote server, a copy of the updated replica partner vector table and object updates to the local server;

improving, by the local server, upon receiving the updated replica partner vector from the remote server, update sequence numbers or timestamps in the partner vector table of the local server; and

calculating, by the local server, a difference between the timestamp for each server in the replica partner vector and a current time.

4. A computer-readable medium as defined in claim 3, further comprising:

transmitting a copy of the remote server'"'"'s replica partner vector table to the local server;

comparing the remote server'"'"'s replica partner vector table with the local server'"'"'s replica partner vector table; and

updating the replica partner vector table received from the remote server; and

transmitting a copy of the updated replica partner vector table and object updates to the remote server.

5. A method for monitoring replication latency in a computer system having at least a first server and a second server that periodically replicate object updates with one another, the method comprising:

Maintaining, on the first server, a replica partner vector table that includes a timestamp indicating a most recent successful replication attempt with the second server; and

calculating a difference between the timestamp in the replica partner vector table and a current time;

comparing the difference to a maximum allowable latency time period; and

generating an alert if the difference is greater than the maximum allowable latency time.

6. A method as defined in claim 5, wherein the alert comprises an email message sent to a network administrator.

7. A method as defined in claim 5, wherein the alert comprises an email message sent to a user in the computer system.

8. A method as defined in claim 5, wherein the alert comprises a broadcast message to all servers in the computer system.

9. A method as defined in claim 5, further comprising:

maintaining, on the first server, an update sequence number that is increased in response to each successful replication attempt with the second server.

10. A method as defined in claim 9, wherein the act of maintaining the replica partner vector table comprises:

maintaining the update sequence number in the replica partner vector table.

11. A method as defined in claim 5, further comprising:

transmitting, by the first server, a copy of the vector partner replica table to the second server; and

receiving, by the first server, an updated copy of the vector partner replica table and one or more object updates from the second server.

12. A method as defined in claim 11, further comprising:

analyzing, by the second server, the copy of the vector partner replica table to determine whether one or more object updates should be replicated on the first server; and

in response to determining that one or more object updates should be replicated on the first server;

creating, by the second server, the updated copy of the vector partner replica table; and

transmitting, by the second server, the updated copy of the vector partner replica table and the one or more object updates.

13. A method as defined in claim 12, wherein the maintaining act comprises:

maintaining in the vector partner replica table an update sequence number that is increased in response to each successful replication attempt with the second server, wherein the analyzing act comprises;

comparing, by the second server, the update sequence number and the timestamp to determine whether one or more object updates should be replicated on the first server.