METHODS AND SYSTEMS FOR REDUCING A LOAD ON A MULTI-TENANT DATABASE

• US 20100235495A1
• Filed: 09/10/2009
• Published: 09/16/2010
• Est. Priority Date: 09/10/2008
• Status: Abandoned Application

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Petersen, David, Soble, Daniel
• Application Number:
  US12/557,256
• Publication Number:
  US 20100235495A1
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: H04L 63/1458 Denial of Service

Abstract

Mechanisms and methods for reducing a load on a multi-tenant database are provided. These mechanisms and methods for reducing a load on a multi-tenant database can enable a reduction in the computational effort expended to handle login requests with invalid usernames and the computational effort expended to handle valid login requests that occur at a high rate. The ability to provide a reduction in computational effort expended on login requests can enable a providing of a reliable level of resources to users and tenants of the multi-tenant database.

Claims

1. A method for reducing a load on a multi-tenant database system having a plurality of tenants, the method comprising:

receiving, at one or more servers of the multi-tenant database system, a plurality of requests to login from one or more users of a first tenant;

determining, with the multi-tenant database system, a number of login requests received from the one or more users of the first tenant;

comparing the number of login requests to a threshold number; and

determining whether the number of login requests is greater than the threshold number.

2. The method of claim 1, further comprising:

when the number of login requests is greater than the threshold number, altering how a login request from the one or more users of the first tenant is processed by the multi-tenant database system.

3. The method of claim 1, wherein the number of login requests is determined by counting login requests received from the one or more users of the first tenant during a predetermined length of time.

4. The method of claim 3, wherein the predetermined length of time is one hour.

5. The method of claim 3, wherein determining a number of login requests received during the predetermined length of time is performed periodically, and comparing the number of login requests to a threshold number is performed after each predetermined length of time.

6. The method of claim 1, wherein altering how a login request from the one or more users of the first tenant is processed includes:

stopping a processing of login requests from the one or more users of the first tenant.

7. The method of claim 1, wherein altering how a login request from the one or more users of the first tenant is processed includes adding a delay for login requests from the one or more users of the first tenant.

8. The method of claim 1, wherein the plurality of login requests are received at a plurality of servers of the multi-tenant database system, and wherein determining the number of login requests received from the one or more users of the first tenant includes:

determining, with each server, a first number of login requests received by the respective server during a first time interval;

sending, from each server to a central processing system, the first number of login requests received by the respective server; and

calculating a total number of login requests received by all of the servers during a second time interval by summing;

the first numbers of login requests received by each respective server during the first time interval; and

a number of login requests previously received during the second time interval, wherein the second time interval is longer than the first time interval and includes the first time interval,

wherein the total number of login requests is used as the number of login requests received from the one or more users of the first tenant.

9. The method of claim 8, wherein the central processing system stores a history of the first numbers of login requests received by the servers during different first time intervals occurring over a rolling time period that has a length equal to the second time interval.

10. The method of claim 8, further comprising:

sending, from the central processing system, the total number of login requests received by all of the servers during the second time interval to each of the servers,

wherein comparing the number of login requests to a threshold number is performed by each server.

11. The method of claim 8, further comprising:

sending, from the central server to each server, an instruction at to whether each tenant has exceed the threshold number.

12. The method of claim 1, further comprising:

determining whether the number of login requests of each of a plurality of the tenants is greater than a respective threshold number, wherein the value of the respective threshold number is dependent on the tenant.

13. The method of claim 1, further comprising:

determining a number of login requests received from each user of the first tenant and a number of login requests received from all of the users of the first tenant;

determining whether the number of login requests received from a first user of the first tenant is greater than a first threshold number; and

determining whether the number of login requests received from all of the users of the first tenant is greater than a second threshold number, wherein the first threshold number is greater than the second threshold number.

14. The method of claim 1, wherein the threshold number is time-dependent.

15. The method of claim 1, wherein the plurality of login requests are received at a plurality of servers of the multi-tenant database system, the method further comprising:

receiving a username at a local database instance as part of a login request;

performing one lookup in a user table stored on the local database instance to identify whether the username exists;

providing one call to a central storage that stores a global user table to identify whether the username exists;

sending, from the local database instance, a request to a plurality of other database instances to perform a lookup in user tables stored at respective database instances to identify whether the username exists;

if the username is not found, stopping a processing of the login request; and

if the username is found, proceeding with an authentication of a password provided with the username.

16. The method of claim 15, further comprising:

using a bloom filter stored in a memory of a first server to check whether a username is in a list of valid usernames; and

if the username is not is in a list of valid usernames, stopping a processing of the login request.

17. The method of claim 15, further comprising:

sending a failed login notice in response to the stopping of a processing of the login request if the username is not found;

adding a delay to the sending of the failed login notice such that the failed login notice using an invalid username is sent at a same elapsed time that a failed login notice using a valid username but with an incorrect password is sent relative to when a login request is received.

18. A machine-readable medium carrying one or more sequences of instructions for reducing a load on a multi-tenant database system having a plurality of tenants, which instructions, when executed by one or more processors, cause the one or more processors to carry out the steps of:

receiving, at one or more servers of the multi-tenant database system, a plurality of requests to login from one or more users of a first tenant;

determining, with the multi-tenant database system, a number of login requests received from the one or more users of the first tenant;

comparing the number of login requests to a threshold number; and

determining whether the number of login requests is greater than the threshold number.

19. A multi-tenant database system comprising:

one or more servers configured to receive a plurality of requests to login from one or more users of a first tenant;

one or more processors configured to;

determine a number of login requests received from the one or more users of the first tenant;

compare the number of login requests to a threshold number; and

determine whether the number of login requests is greater than the threshold number.