Hewlett Packard Enterprise Development LP

International Business Machines Corporation

Christopher S Murray

Sonya Tyler Long

Gregory B Goslen

Howard Thomas Olnowich

William Russell Belknap

Stephen Gouze Luning

Ashok Raj Saxena

Alex Olson

Buddy F Stansbury

James W Eldridge

Stephen Kay

Long B Sam

Larry William Fitchett

Karl David Schubert

Louise Irene Cleary

Jonathan R Guthrie

Buddy Floyd Stansbury

A system monitors metrics relating to link flapping associated with a physical link in a ring network. Network management traffic is transported via a first VLAN and customer data traffic is transported via a second VLAN. The system maintains a signal fail state for the ring network based on the monitored metrics. If a first condition associated with the first VLAN is detected, the system disallows customer data traffic on the second VLAN. If a second condition associated with the monitored metrics is detected, the system disallows network management traffic on the first VLAN and disallows customer data traffic on the second VLAN, resulting in the signal fail state. If the second condition is no longer met, the system allows network management traffic on the first VLAN. If the first condition is no longer met, the system allows customer data traffic on the second VLAN, resulting in the idle state.

Method to stabilize a ring during link flapping

A system for identifying the source of a loop in a ring network is provided. During operation, the system receives a first message, which indicates the presence of a looped packet, from a network device of the ring network. The system receives a second message from a respective network device with an uplink to an external network. The second message can include traffic information associated with the uplink. Upon determining based on the first message that a local link intended to be disabled is active, the system determines that a loop is present in the ring network. Upon determining that a plurality of uplinks are active based on the second message, the system determines that a loop is present in the uplinks. If a loop is not present in the ring network or in the uplinks, the system determines that a loop is present in the external network.

Identifying source of layer-2 loop in ring network

The system determines a first and a second snapshot of memory usage by processes in a computing system based on a predetermined time interval, wherein a respective snapshot indicates, for a respective process, memory usage corresponding to memory types including private memory and shared memory. The system computes, for processes common to the first and the second snapshots, a likelihood of memory leak for a process and memory type based on: an increase in an amount of heap memory and private memory; an increase in an amount of shared memory; and an increase in a total amount of private and shared memory. The system compares the computed likelihood for the process with a predetermined threshold. Responsive to determining that the computed likelihood for the process exceeds the predetermined threshold, the system performs a corrective action to address memory leak associated with the process.

Predicting and mitigating memory leakage in a computer system

Examples described herein relate to a system including a first management system having a primary memory including a free memory, a used memory, and a loosely reserved memory, where the loosely reserved memory comprises cache memory having a reclaimable memory; and a processing resource coupled to the primary memory. The processing resource may monitor an amount of the used memory and an amount of an available memory during runtime of the first management system. Further, the processing resource may enable a synchronized reboot of the first management system if the amount of the used memory is greater than a memory exhaustion first threshold or the amount of the available memory is less than a memory exhaustion second threshold, wherein the memory exhaustion first threshold and the memory exhaustion second threshold are determined based on usage of the reclaimable memory and a number of major page faults.

Managing synchronized reboot of a system

The system determines data which indicate memory usage by a process and information related to a number of instances of the process. The system determines for the process: a first average amount of memory; a first standard deviation; a first value indicating a difference between a current amount of memory used by the process and the first average amount; a second average number of process instances; a second standard deviation; and a second value indicating a difference between a current number of process instances and the second average number. The system performs a corrective action to address a memory spike associated with the process responsive to comparing at least one of: the current amount of memory used by the process with a watermark; the first value with the first standard deviation scaled; the current number of instances with a watermark; and the second value with the second standard deviation scaled.

Detecting spikes in memory usage in a computer system

Techniques for uplink connectivity determination are disclosed. In an example, a Frame Link Module (FLM) in a frame, belonging to a group of frames connected in a ring network, may generate an uplink discovery packet. The FLM may determine, based on a Link Layer Discovery Protocol (LLDP) packet received by the standby uplink from a customer network accessing the ring that the standby uplink has a link to the customer network. The FLM may forward the uplink discovery packet to the standby uplink via a Peripheral Component Interconnect (PCI) interface. The FLM may send the uplink discovery packet to the customer network through the standby uplink directed to an owner FLM. The owner FLM may monitor receipt of the uplink discovery packet from the customer network through a current active uplink and on successful receipt may determine that the standby uplink and switches in the customer network are correctly configured.

Uplink connectivity in ring networks

A method and system are disclosed to enable an FLM designated as an owner, in a group of frames having redundant FLMs per frame, to ensure that an unresponsive 'active' uplink is disabled, prior to allowing another uplink in the frame to become the active uplink. A group of frames, configured in a ring network topology, may only have a single uplink at a time. Otherwise, two or more concurrent active uplinks may cause a network loop that may result in total network failure (e.g., network crash). The disclosed method utilizes a coordination between an FLM owner on the ring and the neighbor FLM, in the same frame as the non-responsive FLM, to disable the active uplink, prior to transitioning to a new active uplink for the group of frames.

Method to disable or reboot unresponsive device with active uplink in a ring network

One embodiment can provide a method and system for monitoring network traffic. During operation, the system may receive a request to monitor packets ingressing or egressing a device. The request can indicate a set of parameters for monitoring the packets. In response to receiving the request, the system may configure a network switch coupled to the device to replicate the packets ingressing or egressing the device. The system can then encapsulate the replicated packets, based on the set of parameters, with a destination address associated with the request, thereby allowing the packets to be analyzed remotely.

System and method for monitoring ingress/egress packets at a network device

Techniques for detecting misconfiguration of an uplink or an external network device connected with the uplink is disclosed. In an aspect, a Frame Link Module (FLM) in a frame belonging to a group of frames connected in a ring network, may determine that an uplink of the FLM is configured in a management mode. The FLM may determine, based on a system description associated with a Link Layer Discovery Protocol (LLDP) packet, whether the uplink is connected to a network device in an external network. Based on Virtual Local Area Network (VLAN) traffic being dropped, it may be determined that VLAN configuration of one of the uplink and the network device is incompatible for operation in the management mode. The uplink with incompatible VLAN configuration or the uplink connected with a network switch having incompatible VLAN configuration may be removed from a group of uplinks available for selection as an active uplink for the frames.

Misconfigured uplink identification

Systems and methods address automated ring owner selection for a ring topology network. A new ring owner may be selected based on failure of a current ring owner. The new ring owner may assume a ring ownership role by a) detecting a missing Ownership Select Message (OSM); b) detecting missing Continuity Check Protocol (CCP) messages; and c) determining that the current ring owner is an adjacent network module connected via an external link of a management network. A signal fail (SF) condition may also be present. Upon detection of adequate role changing conditions, a new ring owner may assume a ring ownership role, transmit an OSM and possibly select a new active uplink.

Rapid owner selection

An example communications device may include communication ports and processing circuitry. The communications device may, when the communication device and peer network nodes are connected in a ring topology to form an access network, detect that a loop exists between the access network and a customer network. In response to detecting the loop, the communication device may automatically block any uplink to the customer network that the communication device may have, and send a message to all of the peer network nodes instructing them to block any uplinks to the customer network they may have.

Responses to loops in networks having a ring topology

Remote configuration of network interface cards (NICs) on appliances of a scaleable compute resource such as a frame-based system is disclosed. Frames may include a frame link module (FLM). Remote configuration of NICs may allow for multiple networks to be maintained in physical or logical isolation from each other. For example, a management data network may be maintained independently of an application data network. An FLM CPU may detect an appliance, validate compatibility for the appliance, a midplane, a PHY connection, etc. Commands from the FLM CPU may configure the independent networks. Independent networks may provide redundancy and segregation by data type. A controller area network (CAN) bus may deliver configuration commands to a NIC of an attached appliance. Air gap equivalent isolation of networks based on type of network may be achieved while maintaining redundancy of networks to address potential failure of individual components.

Remote network interface card management

Examples herein involve dividing a ring protection network into multiple ring protection networks. Examples include detecting a disconnection in a ring protection network, determining whether an owner of the ring protection network is in a new ring protection network divided from the ring protection network, and establishing an owner of the new ring protection network based on whether the owner of the ring protection network is in the new ring protection network.

Growing community of inventors

Boynton Beach, FL, United States of America

Christopher S Murray