Techniques associated with server transaction latency information

Patent metadata

Field	Value
Publication number	US-10355959-B2
Application number	US-201615357415-A
Country	US
Kind code	B2
Filing date	Nov 21, 2016
Priority date	Dec 18, 2012
Publication date	Jul 16, 2019
Grant date	Jul 16, 2019

Abstract

Examples are disclosed for determining or using server transaction latency information. In some examples, a network input/output device coupled to a server may be capable of time stamping information related to ingress request and egress response packets for a transaction. For these examples, elements of the server may be capable of determining transaction latency values based on the time stamped information. The determined transaction latency values may be used to monitor or manage operating characteristics of the server to include an amount of power provided to the server or an ability of the server to support one or more virtual servers. Other examples are described and claimed.

First claim

What is claimed is:

1. A method comprising:
determining, at management logic for a data center, an average channel transaction latency value for a first network communication channel over a first time period, the first network communication channel coupling a first server to a second server, the average channel transaction latency value determined based on a first average transaction latency value for a first set of transactions observed at the first server and a second average transaction latency value for the first set of transactions observed at the second server;
determining an average server transaction latency value for the first server over the first time period, the average server transaction latency value determined based on a third average transaction latency value for incoming requests received at the first server, a fourth average transaction latency value for outgoing requests to one or more servers to include the second server and a total number of the incoming requests received at the first server over the first time period;
generating a topological graph showing the average channel transaction latency for the first network communication channel and the average server transaction latency value for the first server; and
updating the topological graph following a second time period, the updated topological graph to show an updated average channel transaction latency value for the first network communication channel determined over the second time period and an updated average server transaction latency value for the first server also determined over the second time period.

2. The method of claim 1 , comprising determining the average channel transaction latency value for the first network communication channel based on a difference between the first average transaction latency value and the second average transaction latency value.

3. The method of claim 1 , comprising determining the average server transaction latency value for the first server based on a difference between the third average transaction latency value and the fourth average transaction latency value and then dividing the difference by the total number of the incoming requests received at the first server over the first time period.

4. The method of claim 1 , comprising the first server and the second server separately including one or more network input/output devices to time stamp both ingress request packets and egress response packets and gather portions of separately time stamped ingress request and egress response packets to enable the first and second servers to identify separate transactions and to determine separate transaction latency values for the identified separate transactions.

5. The method of claim 4 , comprising the ingress request and the egress response packets associated with the identified separate transactions having transmission control protocol (TCP) and Internet protocol (IP) headers, the gathered portions of the separately time stamped ingress request and egress response packets to include information from TCP and IP headers for the ingress request and egress response packets associated with the identified separate transactions, the information to include a source IP address, a destination IP address, a source port or a destination port.

6. A method comprising:
receiving, at management logic for a data center, transaction latency values from a plurality of physical servers, each physical server capable of supporting one or more virtual servers, the transaction latency values received responsive to a polling request from the management logic;
determining available capacity of each physical server from among the plurality of physical servers to support one or more additional virtual servers based, at least in part, on the received transaction latency values; and
assigning a new virtual server to a first physical server from among the plurality of physical servers based on the first physical server having a highest determined capacity to support one or more additional virtual servers as compared to other physical servers from among the plurality of physical servers, the plurality of physical servers separately including a network input/output device to time stamp both ingress request packets and egress response packets for supported virtual servers, the network input/output device also to gather portions of separately time stamped ingress request and egress response packets to enable the plurality of physical servers to identify separate transactions associated with supported virtual servers and to also enable the plurality of physical servers to determine separate transaction latency values for the identified separate transactions associated with respective supported virtual servers.

7. The method of claim 6 , comprising:
determining that a second physical server from among the plurality of physical servers lacks capacity to support an existing first virtual server based, at least in part, on the received transaction latency values; and
migrating the existing first virtual server to a third physical server from among the plurality of physical servers based on the third physical server having available capacity to support the existing first virtual server.

8. The method of claim 6 , comprising the ingress request and the egress response packets associated with the identified separate transactions having transmission control protocol (TCP) and Internet protocol (IP) headers, the gathered portions of the separately time stamped ingress request and egress response packets to include information from TCP and IP headers for the ingress request and egress response packets associated with the identified separate transactions, the information to include a source IP address, a destination IP address, a source port or a destination port.

9. At least one non-transitory machine readable medium comprising a plurality of instructions that in response to being executed on a server cause the server to:
determine an average channel transaction latency value for a first network communication channel over a first time period, the first network communication channel coupling a first server to a second server, the average channel transaction latency value determined based on a first average transaction latency value for a first set of transactions observed at the first server and a second average transaction latency value for the first set of transactions observed at the second server;
determine an average server transaction latency value for the first server over the first time period, the average server transaction latency value determined based on a third average transaction latency value for incoming requests received at the first server, a fourth average transaction latency value for outgoing requests to one or more servers to include the second server and a total number of the incoming requests received at the first server over the first time period;
generate a topological graph showing the average channel transaction latency for the first network communication channel and the average server transaction latency value for the first server; and
update the topological graph following a second time period, the updated topological graph to show an updated average channel transaction latency value for the first network communication channel determined over the second time period and an updated average server transaction latency value for the first server also determined over the second time period.

10. The at least one non-transitory machine readable medium of claim 9 , comprising the instructions to also cause the server to determine the average channel transaction latency value for the first network communication channel based on a differ

Assignees

• Intel Corp

Inventors

• Deval Manasi
• Daubert Jim
• Mann Eric K
• Li Cong
• Rajappa Muralidhar Murali
• Chagam Reddy Anjaneya Reddy
• Wescott David
• Nagappan Ramkumar
• Kanjo Raed

Classifications

• H04L41/12

  Discovery or management of network topologies · CPC title

• H04L69/28

  Timers or timing mechanisms used in protocols · CPC title

• H04L67/1029 Primary

  using data related to the state of servers by a load balancer · CPC title

• H04L49/354

  for supporting virtual local area networks [VLAN] · CPC title

• H04L69/22

  Parsing or analysis of headers · CPC title