US 8,527,650 B2Granted

Creating a checkpoint for modules on a communications stream

AssigneeInternational Business Machines Corporation

InventorsDavid Jones Craft, Ranadip Das, Rajeev Mishra, Lance Warren Russell

Life of this patent

Estimating the term…

Loading patent…

Abstract

A computer implemented method, computer program product, and system for creating a checkpoint of a stream. A stream checkpoint request to create the checkpoint of the stream is received, wherein the stream is used by a process as a communications path, and wherein the communications path is modified by a set of modules. In response to identifying the identity of each module in the set of modules, the identity of each module in the set of modules is stored in the checkpoint. In response to identifying an order of the set of modules, the order of the set of modules is stored in the checkpoint. In response to sending a stream checkpoint message to each module in the set of modules, module data is received from each module in the set of modules to form received module data. The received module data is stored in the checkpoint.

Claims20

Dependent claims

A method comprising:

responsive to a request to move a software partition from a departure server to an arrival server, freezing one or more processes of the software partition and freezing one or more communications of the one or more processes, wherein the one or more processes use a communication path for the one or more communications between the departure server and the arrival server;
creating a set of checkpoint information for the software partition including state information for the one or more processes and state information for the one or more communications in the communication path;
using the communication path to send the checkpoint information to the arrival server for resuming the one or more processes and the one or more communications in the arrival server;
responsive to receiving a checkpoint request, creating, by a first processor on the departure server, a first checkpoint data in the set of checkpoints for a stream of the one or more communications, wherein the first checkpoint data checkpoint comprises state information for a set of processes executing on the departure server;
responsive to receiving the checkpoint request, creating, by the first processor on the departure server, a second checkpoint data in the set of checkpoints for a set of modules used to send the stream to a transport layer, wherein the first checkpoint data and the second checkpoint data form the checkpoint, wherein the step of creating the second checkpoint data for the set of modules further comprises responsive to identifying an identity of each module in the set of modules, storing the identity of the each module in the set of modules in the second checkpoint data;
responsive to identifying an order of the set of modules, storing the order of the set of modules in the second checkpoint data;
responsive to sending a stream checkpoint message to the each module in the set of modules, receiving module data from the each module in the set of modules to form received module data;
storing the received module data in the second checkpoint data;
restarting, by a second processor on the arrival server, the stream after the first checkpoint and the second checkpoint have been created;
and
restarting, by the second processor on the arrival server, the set of modules after the first checkpoint and the second checkpoint have been created;
wherein the communication path is to a transport layer in the departure server and the transport layer further communicates over a network with at least one other transport layer in the departure server.

9 more elements

A computer program product comprising:

one or more non-transitory computer-readable mediums;
computer usable program code, stored in at least one non-transitory computer-readable mediums, responsive to a request to move a software partition from a departure server to an arrival server, for freezing one or more processes of the software partition and freezing one or more communications of the one or more processes, wherein the one or more processes use a communication path for the one or more communications between the departure server and the arrival server;
computer usable program code, stored in at least one of the non-transitory computer-readable mediums, for creating a set of checkpoint information for the software partition including state information for the one or more processes and state information for the one or more communications in the communication path;
computer usable program code, stored in at least one of the non-transitory computer-readable mediums, for using the communication path to send the checkpoint information to the arrival server for resuming the one or more processes and the one or more communications in the arrival server;
computer usable program code, stored in at least one the non-transitory computer-readable mediums, for creating a first checkpoint data in the first set of checkpoints for a stream of the one or more communications on a departure server in response to receiving a checkpoint request, wherein the first checkpoint data checkpoint comprises state information for a set of processes executing on the departure server;
computer usable program code, stored in at least one of the non-transitory computer-readable mediums, for creating a second checkpoint data in the set of checkpoints on the departure server for a set of modules used to send the stream to a transport layer in response to receiving the checkpoint request, wherein the first checkpoint data and the second checkpoint data form the checkpoint, wherein the computer usable program code for creating the second checkpoint data for the set of modules further comprises responsive to identifying an identity of each module in the set of modules, computer usable program code for storing the identity of the each module in the set of modules in the second checkpoint data responsive to identifying an order of the set of modules, computer usable program code for storing the order of the set of modules in the second checkpoint data responsive to sending a stream checkpoint message to the each module in the set of modules, computer usable program code for receiving module data from the each module in the set of modules to form received module data storing the received module data in the second checkpoint data;
computer usable program code, stored in the non-transitory computer-readable medium, for restarting the stream on the arrival server after the first checkpoint and the second checkpoint have been created;
and
computer usable program code, stored in at least one of the non-transitory computer-readable mediums, for restarting the set of modules on the arrival server after the first checkpoint and the second checkpoint have been created;
wherein the communication path is to the transport layer in the departure server and the transport layer further communicates over a network with at least one other transport layer in the departure server.

7 more elements

A data processing system comprising:

a departure server having one or more first processing units, one or more first non-transitory computer readable mediums and one or more first computer readable memories connected to a first bus;
a first communications unit connected to the first bus;
an arrival server having one or more second processing units, one or more second non-transitory computer readable mediums and one or more second computer readable memories connected to a second bus;
a second communications unit connected to the second bus;
and
computer usable program code, stored in the first non-transitory computer-readable medium for execution by at least one processing unit via at least one memory, responsive to a request to move a software partition from a departure server to an arrival server, for freezing one or more processes of the software partition and freezing one or more communications of the one or more processes, wherein the one or more processes use a communication path for the one or more communications between the departure server and the arrival server;
computer usable program code, stored in the first non-transitory computer-readable medium, for creating a set of checkpoint information for the software partition including state information for the one or more processes and state information for the one or more communications in the communication path;
computer usable program code, stored in the first non-transitory computer-readable medium, for using the communication path to send the checkpoint information to the arrival server for resuming the one or more processes and the one of more communications in the arrival server;
computer usable program code, stored in at least one of the first non-transitory computer-readable mediums for execution by at least one first processing unit via at least one first memory for creating a first checkpoint data in the set of checkpoints on a departure server for a stream of the one or more communications in response to receiving a checkpoint request, wherein the first checkpoint data checkpoint comprises state information for a set of processes executing on the departure server;
create a second checkpoint data in the set of checkpoints on the departure server for a set of modules used to send the stream to a transport layer in response to receiving the checkpoint request for the stream, wherein the first checkpoint data and the second checkpoint data form the checkpoint, wherein creating the second checkpoint data for the set of modules further comprises responsive to identifying an identity of each module in the set of modules, storing the identity of the each module in the set of modules in the second checkpoint data responsive to identifying an order of the set of modules, storing the order of the set of modules in the second checkpoint data responsive to sending a stream checkpoint message to the each module in the set of modules, receiving module data from the each module in the set of modules to form received module data and storing the received module data in the second checkpoint data;
and
computer usable program code, stored in at least one of the second non-transitory computer-readable mediums for execution by at least one second processing unit via at least one second memory for restarting the stream on the arrival server after the first checkpoint and the second checkpoint have been created;
and
computer usable program code, stored in at least one of the second non-transitory computer-readable mediums for execution by at least one second processing unit via at least one second memory for restarting the set of modules on the arrival server after the first checkpoint and the second checkpoint have been created;
wherein the communication path is to the transport layer in the departure server and the transport layer further communicates over a network with at least one other transport layer in the departure server.

12 more elements

Dependencies are inferred from the claim text.