Browse Prior Art Database

A method to deploy location-based servers naming convention in distributed system

IP.com Disclosure Number: IPCOM000244206D
Publication Date: 2015-Nov-24
Document File: 3 page(s) / 61K

Publishing Venue

The IP.com Prior Art Database

Abstract

This disclosure describes a method to deploy the location based servers naming convention in the distributed system, usually a bunch of servers or a cluster/cloud in datacenter. Using this method, the system admin could predefine the location-based naming convention of the servers based on the servers' inventory information, and then the system could automatically deploy the location-based naming convention to the servers, this could significantly save the time of manually naming the servers. The brief procedure of the automatic process is: 1. Predefine the servers naming policy in the system management tool head node, the policy is based on physical locations. 2. The servers send the inventory information to the system management tool head node. 3. The head node will calculate the system configuration for the servers, using the naming policy and the inventory information. 4. Based on the result from step #3, push the configuration to the servers through in-band or out-band channels.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 42% of the total text.

Page 01 of 3

A method to deploy location

A method to deploy location-

--based servers naming convention in distributed system

based servers naming convention in distributed system

With the increasing requirements with the data processing and cloud popularity, the datacenters have become a very natural way for the companies to consolidate and simplify their IT infrastructure. Such data centers can include a large number of individual servers systems, the datacenters usually have some kind of uniformed system management portfolio, there are a lot of scenarios in which the servers are anonymous units in the system management portfolio, the system management tool only knows this is a server system on the network and in the datacenter, with arbitrary or random hostnames, ip addresses and service processor configuration, but the system management tool does not know which is which, it brings in difficulties for the system maintainability, for example, if the system admin gets a hardware service events and needs to physically check the server system or replace some FRUs(Field Replacement Units) on a node with name, say node0035, but the system admin does not know which physical machine this node is. Or, the users report that the service processor 172.20.3.58 is not responsive, the system admin also needs to know which is which.

Here are some scenarios in which the servers are anonymous units for the system management tool:

1) For the new servers in datacenter, if the servers have pre-installed operating system, when the servers are powered on, a common practice is to use DHCP servers to assign dynamic ip addresses to the operating systems on the servers, when the servers are up and running with dynamic DHCP addresses, the system management tool could discover the servers through the dynamic operating system ip address, then these new servers are anonymous units for the system management tool.

2) In the bare metal case, the servers do not have pre-installed operating systems, there are a lot of ways to automatically deploy operating system on the servers through network booting protocols. Some prior arts indicate some methods, for example, https://www.google.com/patents/EP2172840A1, http://blogs.technet.com/b/adiyq/archive/2014/01/30/end-to-end-bare-metal-provisioning-with-scvmm-2012-sp1-r2-part-1.aspx

and

                       . In the servers deployment phase, and when the servers are deployed with operating system, they usually uses dynamic hostname and ip addresses so they are anonymous units.

3) In the cloud environment, the cloud software usually uses arbitrary node names, if another infrastructure management tool is being used in the cloud environment, the cloud instances nodes are anonymous units.

To naming the anonymous server systems, an obvious method is to naming the servers based on the physical location, for example, the r35u24 means the node in unit 24 of rack 35. We did not find any existing method to automatically naming these anonymous server systems base...