International Business Machines Corporation

David Zhiwei Dai

Dwip N Banerjee

Khalid Filali-Adib

Frederic Barrat

Vasu Vallabhaneni

Patrick Tam Vo

Lilian Sylvia Fernandes

Rashmi Narasimhan

Richard Perry Youngman

Ine Youngman, Legal Representative

In a routing table of a first system, a routing table entry for a first application is created by a controller application executing in a controller system, the first application intended for relocation, a second application in the set of applications not intended for relocation. At the second system responsive to commencing execution of the first application in the second system at a second time, a mapping entry is created by a controller application, the mapping entry causing the forwarded packet to be delivered to the first application at the second system. Responsive to a notification that the first application is executing in the second system, the routing table entry of the first system is modified by the controller application, the modified entry causing a packet addressed to the first application to be forwarded from the first system to the second system.

Live application and kernel migration using routing table entries

Technical solutions are described for performing a live update of an operating system. An example method includes cloning an original root volume group associated with an operating system instance executing in a first logical partition to generate a cloned root volume group for booting a second logical partition. The method further includes applying the update to the cloned root volume group, and booting the second logical partition. The method further includes migrating an application instance, from the original root volume group to the second logical partition. The migration includes creating a network tunnel between the first and second logical partitions, and restarting the application instance on the second logical partition. The migration further includes forwarding data directed to the application instance on the first logical partition to the application instance on the second logical partition, and assigning a network address of the first logical partition to the second logical partition.

Live kernel updating using progressive checkpointing and network tunnels

The present disclosure relates to maintaining network services during kernel update operations. According to one embodiment, a server generates a second computing system executing a second system kernel. The second system kernel is updated in relation to a first system kernel executing on a first computing system. The server pauses network services operating on the first computing system. The server generates a snapshot of the network services operating on the first computing system. While the network services are paused on the first computing system, the server caches data packets received from client devices and destined for network services operating on the first system, migrates the network services to the second computing system, and restores the network service snapshot to the second computing system. Upon completing the migration, the server sends the cached data packets to the one or more network services operating on the second computing system.

User datagram protocol (UDP) application handling during live kernel update

A method for handling transmission control protocol (TCP) data between networked computer nodes during blackout periods associated with live kernel updating is provided. The method includes initiating a live kernel update on a first node, transmitting, prior to performing a network blocking procedure associated with the live kernel update, a zero window update to one or more peer nodes, generating a surrogate partition associated with the first node, copying network states from an original partition associated with the first node to the surrogate partition, restoring the network states on the surrogate partition, and transmitting another window update to the one or more peer nodes.

Transmission control protocol (TPC) data handling

Transmission control protocol (TCP) data handling

A mechanism is provided to enhance the scalability of network caching capabilities. All network client applications running on a partition in a virtualized environment are configured to query information from a single virtual input/output server (VIOS) network caching daemon. Thus, the illustrative embodiments provide a 1:n model where a VIOS partition has a network caching daemon, and each of the n partitions uses the network caching daemon of the VIOS partition. The mechanism of the illustrative embodiments only requires the system administrator to control one copy of the local files on the VIOS server. The system administrator does not need to monitor all of these local files on each individual partition. The system administrator consolidates the entries used by the different individual partitions into one single file on the VIOS server side.

Enhancing the scalability of network caching capability in virtualized environment

A method, system, and computer-usable medium for enhancing environment variable and process option invocation in a data processing system. In a preferred embodiment of the present invention, in response to receiving a request for a system startup, an operating system loads a centralized environment file. In response to receiving a request for a user login, the operating system searches for a user environment file. In response to determining that the user environment file conflicts with the centralized environment file, the operating system overrides the user environment file with settings from the centralized environment file.

System and method for enhancing environment variable and process option invocation

A method in a data processing system for routing a packet to a target data processing system. With the method and apparatus, a global address for the target data processing system is placed in the packet. A hop-by-hop option may then be set in the packet, which contains a media access control address for the final destination. This option causes a router for a subnet on which the target data processing system is located to send the packet to the target data processing system using the media access control address without performing a link level address discovery operation.

Growing community of inventors

Austin, TX, United States of America

David Zhiwei Dai