GB2370140A - A delay accounting method for computer system response time improvement - Google Patents

A delay accounting method for computer system response time improvement Download PDF

Info

Publication number
GB2370140A
GB2370140A GB0120564A GB0120564A GB2370140A GB 2370140 A GB2370140 A GB 2370140A GB 0120564 A GB0120564 A GB 0120564A GB 0120564 A GB0120564 A GB 0120564A GB 2370140 A GB2370140 A GB 2370140A
Authority
GB
United Kingdom
Prior art keywords
time
service
user
resource
summing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
GB0120564A
Other versions
GB0120564D0 (en
Inventor
Jr Thomas Edwin Turicchi
Doug Grumann
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
HP Inc
Original Assignee
Hewlett Packard Co
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hewlett Packard Co filed Critical Hewlett Packard Co
Publication of GB0120564D0 publication Critical patent/GB0120564D0/en
Publication of GB2370140A publication Critical patent/GB2370140A/en
Withdrawn legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/30Monitoring
    • G06F11/34Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
    • G06F11/3419Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment by assessing time
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/30Monitoring
    • G06F11/34Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3438Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment monitoring of user actions

Abstract

A method for determining the system resource parameters <B>[220]</B> to change for specified users <B>[215]</B> or groups of users <B>[215]</B> in order to improve system performance as defined by response times <B>[115]</B>. Methods are disclosed identifying system resources <B>[205]</B> whose parameters <B>[220]</B> are in need of adjustment in order to improve response time <B>[115]</B> for service by those resources <B>[220]</B>. In a representative embodiment, measured results of response time <B>[115]</B> for service provided by various system resources <B>[205]</B> is aggregated for a selected user <B>[215]</B> or group of users <B>[215]</B>. Averages of such aggregations indicate which resources <B>[220]</B> consume excessive amounts of time for processes <B>[210]</B> used by the user <B>[215]</B> or group of users <B>[215]</B>. Beneficial changes in values of system resources <B>[205]</B> related parameters <B>[220]</B> are then implemented.

Description

A DELAY ACCOUNTING METHOD FOR
COMPUTER SYSTEM RESPONSE TIME
IMPROVEMENT
FIELD OF THE INVENTION
The present invention relates generally to computer systems and, more particularly, to the adjustment of parameters to values that result in improved system 10 performance as reflected in the service level provided to various users.
BACKGROUND OF THE INVENTION
15 Modern computing systems generally have numerous processes, often belonging to different users, running on the computer system at the same time. In order to improve system performance, it is often desirable to provide a different level of service for different processes. Administrators of modem computing systems generally have various parameters available that can be adjusted or configured to values that result in improved 20 performance of the system. Size of swap space, memory dedicated to buffer-cache, maximum number of processes permitted, and maximum application data size are examples of such parameters that could be so adjusted.
One method of determining which system parameters can be used to manipulate service-level is by educated guesswork and experimentation by the operator. The 25 disadvantages of this method are (1) it is time intensive for the computer operator, (2) its effectiveness is limited by the competency of the operator, and (3) in order to be effective under changing system loads and circumstances, it must be done continuously.
An alternative and more common method is to only change a very limited number of parameters, usually those that the operator is most familiar with. This has the
disadvantage that it is often not as effective as one would like.
Thus, there is a need for a method by which a computer system can be more effectively and efficiently improved. In particular, there is a need for a method to automatically and continually adjust system parameters to obtain improved system 5 performance, even as the workload on the system changes.
SUMMARY OF THE INVENTION
The present patent document relates to a novel method for improving computer system performance by the adjustment of computer system parameters. This method can 5 make appropriate adjustments even as the workload on the system changes. Previous methods for improving system performance have relied upon speculation as to the optimal system configuration or by controlling a limited number of system parameters.
Modern computer systems typically service a number of users and each of these users will have at various times one or more processes operating on the computer. Each 10 process may need a variety of system resources often including the same resources needed by another process or processes. For the very restricted case of only one process requesting service by a system resource, the time delay for servicing the request by the process is the time which the resource needs in order to perform that request. For multi user computer systems, an additional increment of time, the time delay results from the 15 process having to wait until the resource is available to the process. The sum of these two times is the total time which the process will experience in requesting service by the resource. For a single user having a single active process this sum is also referred to as the response tinge, i.e., the time in which the user receives the service which he requested r from the process. More than one process by the user will require a summation over all 20 the processes which the user has activated.
When running various processes by the same or different users multiple workloads on the same system, or when attempting to give different users of the same workload different service levels, one needs the capability to (a) group processes according to the users or workload that they are performing, and (b) use the controls 25 (knobs) provided by tl e system to allocate resources to the different groups of processes in order to control the service level that the end-users of those workloads receives.
Modern computing systems have a variety of controls available, but it is often very difficult for the system operator to determine how to set or change the "knob" settings in order to manipulate the service level that the end-user receives. Representative 30 embodiments disclosed herein monitor the running system and advise the operator which
"knobs" can be used most effectively to manipulate the service-level of each workload.
In order to remove double counting of wait times, during those times that a user has more than one process waiting for service by a resource, the method counts that part of the wait time only once. The summations assume a single microprocessor. In cases 5 where there are multiple microprocessors, some of the wait times, as well as processing times, occur in parallel and the response times at various levels will be less than or equal to that indicated by the equations.
In a representative embodiment, the time delay experienced by all processes for efforts done on behalf of a particular user during a request for service is aggregated. The 10 time delay is further aggregated and averaged on behalf of that user over multiple requests for service. This aggregation could also be done for a group of related users.
Decisions as to changing or not changing system parameter values for selected system resources is then notate typically based on the results of the delay times obtained.
Those system resources that have the longer delays are easily identified and appropriate 15 changes made in order to improve the response time for the user or user groups of interest. Representative embodiments, as described in the present patent document, are advantageous in that they take very little time or effort on the part of the system operator to arrive at an improved system configuration. Additionally, as the usage characteristics 20 and load on the system change, this method finds and recommends new improvements to the system, without requiring the operator to continuously monitor and redo previous adjustments that were performed by hand. Another benefit is that this mechanism can be used on a subset of the parameters available on a system. Thus, the operator may choose to use this method to look for improvements in obscure or less understood parameters, 25 and make the macroscopic adjustments himself.
The operator is provided much greater visibility into how to best manipulate the service levels experienced by a user or group of users. The method is time efficient and the actions taken by the operator can be effective in response time improvement.
Other aspects and advantages of the present invention will become apparent from 30 the following detailed description, taken in conjunction with the accompanying drawings,
illustrating by way of example the principles of the invention.
s
BRIEF DESCRIPTION OF THE DRAWINGS
The accompanying drawings provide visual representations which will be used to more fully describe the invention and can be used by those skilled in the art to better 5 understand it and its inherent advantages. In these drawings, like reference numerals identify corresponding elements and: Figure 1 is a drawing of an equation for a time delay experienced by a process requesting service by a resource as described in various representative embodiments of the present patent document.
10 Figure 2 is a diagram of a user-process-resource interaction as described in various representative embodiments of the present patent document.
Figure 3 is a flow chart of a process for accumulating data for a delay accounting method as described in various representative embodiments of the present patent document.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
1. Introduction:
As shown in the drawings for purposes of illustration, the present patent 5 document relates to a novel method for improving computer system performance by the adjustment of computer system parameters. This method can make appropriate adjustments even as the workload on the system changes. Previous methods for improving system performance have relied upon speculation as to the optional system configuration or by controlling a limited number of system parameters. In the following 10 detailed description and in the several figures of the drawings, like elements are identified
with like reference numerals.
When running various processes by the same or different users multiple workloads on the same system, or when attempting to give different users of the same workload different service levels, one requires the capability to (a) group processes 15 according to the users or workload that they're performing, and (b) use the controls (knobs) provided by the system to allocate resources to the different groups of processes in order to control the service level that the end-users of those workloads receives.
Modern computing systems have a variety of controls available, but it is often very difficult for the system operator to determine how to set or change the "knob" settings in 20 order to manipulate the service level that the end-user receives. Representative embodiments disclosed herein monitor the ruMing system and advise the operator which "knobs" can be used most effectively to manipulate the service-level of each workload.
2. The Computer System: 25 Figure 1 is a drawing of an equation for a time delay 115 experienced by a process requesting service by a resource as described in various representative embodiments of the present patent document. Modern computer systems typically service a number of users and each of these users will have at various times one or more processes operating on the computer. Each process may need a variety of system 30 resources often including the same resources needed by another process or processes. For
the very restricted case of only one process requesting service by a system resource, the time delay 115 for servicing the request by the process is the time 105 which the resource needs to perform that request. For multi-user computer systems, an additional time delay 110 results from the process having to wait until the resource is available to the process.
5 The sum of these two times is the total time 115 which the process will experience in requesting service by the resource. For a single user having a single active process this sum is also referred to as the response time 115, i.e., the time in which the user receives the service which he requested from the process. More than one process by the user will require a summation over all the processes which the user has activated.
10 Figure 2 is a diagram of a user-process-resource interaction as described in various representative embodiments of the present patent document. The computer system comprises a number of resources 205 only three of which, k-1, k, and k+1 are shown in figure 2. Other resources are indicated as rows of three dots above and below the three resources shown. Service may be requested by the various process 210 operable 15 on the computer system. The computer system further comprises a number of processes 210, only three of which, j-1, j, and j+1 are shown in figure 2. Other processes 210 are indicated as rows of three dots above and below the three processes shown. Various processes 210 are activated at various times by a computer system user 215 identified as the nth user or "user n" in figure 2. Typically the computer system further comprises a 20 number of users 215 not shown in figure 2. While figure 2 implies that any process 210 can access any resource 205, it will be recognized by one of ordinary skill in the art that some resources 205 may not be accessible by particular processes 210. Similarly figure 2 seems to imply that any user 215 can activate any process 210. However, it will be recognized by one of ordinary skill in the art that some processes 210 may not be 25 accessible by particular users 215.
3. Definitions & Equations: The following definitions and equations describe time delays that will be experienced by the various users 215 of the computer system.
Service Level for a user for a given request = Response Time experienced for that request The following symbols represent integers which are used as indices and which identify the item specified.
s n - user n, the nth user of the computer system I = identifying number associated with the ith request for service by a given user 10 j = identifying number associated with Pj, the jth process k = identifying number associated with Rat, the lath resource TA(n,i,Pj,Rk) = Time that the lath resource, Rk, spends in active worl; for the jth process, 15 Pj, during the ith request for service by user n 215 Tw(n i pi,Rk) = Time that the jth process (Pj) spends waiting for the lath resource (Rk) during the ith request for service by user n 215 20 For a single process 210, the jth process 210, activated by the nth user 215, requesting service from a single resource 205, the lath resource 205, during the ith request by the nth user 215, the response time or time delay experienced is Ts(n,i,Pj,R.)- Tw(nji pj,Rk) + TA(n,i,Pj,R) (Eq. l) T otal response time that user n 215 experiences for all processes Pj 210 using resources Rk 205 during the ith request by user n 215 for service.
TR (n,i,Rk) = Ts(n,i,Pj,Rk) = it, Tw(n i pj,Rk) TA(n,i,Pj,Rk) (Eq. 2) j i i
In order to remove double counting of wait times, during those times that user n 215 has more than one process 210 waiting for service by resource Ret, Tw n,i,Pj,R counts that part of the wait time only once. 3 5 Average response time that user n 215 experiences for all processes Pj 210 using resources Rk 205 during multiple, for example a total of I requests, requests by user n 215 for service TAVG RK(n) = I, TRI;(n,i,Rk) / I (Eq 3) Total response time that user n experiences for all processes Pj 210 using all resources Rk 205 during the ith request by user n 215 for service 15 TR(n,i) = i, Ts(n,i,Pj,Rk) i, i, Tw(n,i,Pj,R') + i, i, TA(n,i,Pj,R) (Eq. 4) j k j k j k For a total of I requests by user n 215, the average response time for user n is 20 TAvG(n) = [ TR(n,i)]/I (Eq 5) For a total of N users 215, the average response time for each user is 25 TALL = [ TAVG(n)]/N (Eq. 6) The above summations assume a single microprocessor. In cases where there are multiple microprocessors, some of the wait times, as well as processing times, occur in 30 parallel and the response times at various levels will be less than or equal to that above.
4. Representative embodiments: Figure 3 is a flow chart of a process for accumulating data for a delay accounting method as described in various representative embodiments of the present patent document. In block 305 the user 215, for example user n 215, requests service by process 5 210, for example process j 210. Block 305 then transfers control to block 310.
In block 310, process j 210 requests service of system resource 205, for example resource k 205. Block 310 then transfers control to block 315.
In block 315, resource k 205 records the time process j 210 requested its service, also referred to herein as a first time. Block 315 then transfers control to block 335.
to In block 320 another user 215, for example user n-1 215, requests service by the same process j 210. Blocs; 320 then transfers control to block 325.
In block 325, process j- l 210 requests service of the same system resource k 205, for example resource k 205. Block 325 then transfers control to block 330.
In block 330, resource k 205 records the time process j-1 210 requested its 15 service. Block 330 then transfers control to block 335.
In block 335 resource k 205 selects which of the two competing requests for service that it will service next, for example the request from process j 210. Block 335 then transfers control to block 340.
In block 340 resource k 205 records salient system resource control parameter 20 values for resource k 205. Block 340 then transfers control to block 345.
In block 345 resource k records the time that the system resource k 205 begins service to process j 210, also referred to herein as a second time. Block 345 then transfers control to block 350.
In block 350 system resource k 205 performs the service requested by process j 25 210. Block 350 then transfers control to block 355.
In block 355 resource k records the time that the system resource k 205 ends service to process j 210, also referred to herein as a third time. Block 355 then transfers control to block 360.
In block 360 resource k returns control to process j 210.
30 Tw(n i pi,Rk) which, as defined above, is the time that process j 210 spends
waiting for service by resource Ic 205 during the ith request for service by user n 21 is obtained as the subtraction of the clock time recorded in block 315 (time request for service made) from the clock time recorded in block 345 (time service begun).
TA(n,i,Pj,Rk) which, as defined above, is the time that resource k 205 spends in active 5 work for the process j 210 is obtained as the subtraction of the clock time recorded in block 345 (time service begun) from the clock time recorded in block 3i5 (time service ended). For process j 210, activated by the user n 215, requesting service from resource k 20, during the ith request by user n 215, the response time or time delay experienced is then Ts(n,i,Pj,Rk) as computed in equation 1.
10 In a representative embodiment, the time delay experienced by all processes for efforts done on behalf of a particular user 215, as for example user n 215, during the ith request for service is aggregated as given by equation 2, and further aggregated and averaged on behalf of that user 215, as for example user n 215, during multiple requests for service is aggregated as given by equation 3. This aggregation could also be done for 15 a group of related users 215.
Decisions as to changing or not changing the system parameter value for system resource 205 is then made typically based on the results of equation 3. Those system resources that have the longer delays are easily identified and appropriate changes made in order to improve the response time for the user or user groups of interest. It should be 20 noted that other aggregations and averages are possible as specified in equations 46.
5. Concluding Remarks: Representative embodiments, as described in the present patent document, are advantageous in that they take very little time or effort on the part of the system operator 25 to arrive at an unproved system configuration. Additionally, as the usage characteristics and load on the system change, this method finds and recommends new improvements to the system, without requiring the operator to continuously monitor and redo previous adjustments that were performed by hand. Another benefit is that this mechanism can be used on a subset of the parameters available on a system. Thus, the operator may choose 30 to use this method to look for improvements in obscure or less understood parameters,
and make the macroscopic adjustments himself.
The operator is provided much greater visibility into how to best manipulate the service levels experienced by a user or group of users. The method is time efficient and the actions taken by the operator can be effective in response time improvement.
5 While the present invention has been described in detail in relation to representative embodiments thereof, the described embodiments have been presented by way of example and not by way of limitation. It will be understood by those skilled in the art that various changes may be made in the form and details of the described embodiments resulting in equivalent embodiment that remains within the scope of the 10 appended claims.

Claims (1)

1. A program storage medium readable by a computer, tangibly embodying 2 a software program of instructions executablebythe computer to perform method steps for computer system performance improvement, the 4 method steps comprising: 6 recording a first time 1315], wherein the first time is the computer clock time that a user's 1215] computer process [2101 requests service from a 8 computer system resource [205]; 10 recording value [340] of at least one preselected computer system resource parameter [220]; recording a second time [345], wherein the second time is the computer 14 crock time that the resource [205] initiates service to the computerprocess 1210; recording a third time [355], wherein the third time is the computer clock 18 time that the resource [205] completes the requested service; 20 subtracting the first time [315] from the second time [345] and assigning the result to a wait time [110];
subtracting the second time [345] from the third time [355] and assigning 24 the result to a processing time ll 05]; 26 summing the wait time lllO] and the processing time llO5]; 28 assigning a response time [115] to the result of the method step summing the wait time 11101 and the processing time llO5]; and when a user service time exceeds a preselected value, wherein the user 32 service time is dependent upon the response time 1115] , 34 changing one of the recorded parameter values 1220].
2. A program storage medium as recited in claim 1, providing the user 2 service time is equal to the response time [115].
3. A program storage medium as recited in claim 1, wherein the method 2 steps further comprise: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 1 for at least one additional process 1210] requesting resource [205] service; summing the response times 1115] obtained for the processes 1210], 10 wherein when more than one process [210] is waiting for service, the wait time lllO] is included in only one ofthe response times 1115]; and setting the user service time equal to the result of the method step 14 summing response times 1115].
4. A program storage medium as recited in claim 1, wherein the method 2 steps further comprise: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 1 for at least one additional user [215] resource [205] service request, and for at least one additional process [210] 8 requesting resource [205] service for at least one of the user [215] resource [205] service requests; summing the response times [115] obtained for the processes [210], 12 wherein when more than one process [210] is waiting for service, the wait time [110] is included in only one of the response times 1115]; dividing the results ofthe method step summing the response times [115] 16 by the number of requests activated by the user [215]; and 18 setting the user service time equal to the result of the method step dividing the results ofthe method step summing the response times [115].
5. A program storage medium as recited in claim 1, wherein the method 2 steps further comprise: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 1 for at least one additional process [210] requesting resource [205] service, and for at least one additional resource 8 [205] from which at least one of the processes [210] requests resource [205] service;
10 summing the response times [115] obtained for the processes 1210], wherein when more than one process [210] is waiting for service, the wait 12 time lllO] is included in only one of the response times [115]; and 14 setting the user service time equal to the result of the method step summing response times 1115].
6. A program storage medium as recited in claim 1, wherein the method 2 steps further comprise: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 1 for at least one additional user [215] resource [205] service request, for at least one additional process [210] requesting 8 resource [205] service for at least one of the user [215] resource [205] service requests, and for at least one additional resource 1205] from which 10 at least one of the processes [210] requests resource [205] service for at least one of the user [215] resource service requests; summing the response times [115] obtained for the processes [210], 14 wherein when more than one process [210] is waiting for service, the wait time [110] is included in only one ofthe response times [115]; dividing the results of the method step summing the response times [115] 1 8 by the number of requests by the user [2151; and 20 setting the user service time equal to the result of the method step dividing the results of the method step summing the response times [115].
7. A program storage medium as recited in claim 1, wherein the method
2 steps further comprise: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 1 for at least one additional user [215], for at least one additional user [215] resource 1205] service request by at least 8 one ofthe users [215], for at least one additional process [210] requesting resource [205] service for at least one of the user [215] resource [205] 10 service requests by at least one of the users 1215], and for at least one additional resource [205] from which at least one of the processes 1210] 12 requests resource [205] service for at least one of the user [215] resource service requests by at least one of the users [215]; summing the response times [115] obtained for the processes [2101, 16 wherein when more than one process [210] is waiting for service, the wait time [110] is included in only one of the response times [115]; performing all above method steps for at least one additional user [215]; dividing the results of the method step summing the response times [115] 22 by the number of requests by the users [215]; and 24 setting the user service time equal to the result of the method step dividing the results of the method step summing the response times [115].
8. A computer operable method for computer system performance 2 improvement, comprising the steps of: 4 recording a first time [315], wherein the first time is the computer clock time that a user's [215] computer process [210] requests service from a
6 computer system resource [205]; 8 recording value [340] of at least one preselected computer system resource parameter 1220]; recording a second time 1345], wherein the second time is the computer 12 clock time that the resource 1205] initiates service to the computer process [210];
recording a third time 1355], wherein the third time is the computer clock 16 time that the resource 1205] completes the requested service; 18 subtracting the first time [315] from the second time 1345] and assigning the result to a wait time 1110]; subtracting the second time [345] from the third time 13551 and assigning 22 the result to a processing time [105]; 24 summing the wait time [110] and the processing time [105]; 26 assigning a response time 1115] to the result of the method step summing the wait time 1110] and the processing time 1105]; and when a user service time exceeds a preselected value, wherein the user 30 service time is dependent upon the response time [115], 32 changing one of the recorded parameter values 1220].
9. A computer operable method as recited in claim 8, providing the user 2 service time is equal to the response time 1115].
10. A computer operable method as recited in claim 8, which further 2 comprises the method steps of: 4 except the method step changing one of the recorded parameter values [220] when the user service time exceeds a preselected value, performing 6 the method steps of claim 15 for at least one additional process [210] requesting resource [205] service; summing the response times [115] obtained for the processes 1210], 10 wherein when more than one process [210] is waiting for service, the wait time [110] is included in only one oftheresponse times [115]; and setting the user service time equal to the result of the method step 14 summing response times [115].
!
GB0120564A 2000-08-31 2001-08-23 A delay accounting method for computer system response time improvement Withdrawn GB2370140A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US65251000A 2000-08-31 2000-08-31

Publications (2)

Publication Number Publication Date
GB0120564D0 GB0120564D0 (en) 2001-10-17
GB2370140A true GB2370140A (en) 2002-06-19

Family

ID=24617083

Family Applications (1)

Application Number Title Priority Date Filing Date
GB0120564A Withdrawn GB2370140A (en) 2000-08-31 2001-08-23 A delay accounting method for computer system response time improvement

Country Status (2)

Country Link
JP (1) JP2002123403A (en)
GB (1) GB2370140A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015011487A1 (en) 2013-07-24 2015-01-29 David Mchattie Monitoring the performance of a computer

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1998052122A1 (en) * 1997-05-14 1998-11-19 Compuware Corporation Accurate profile and timing information for multitasking systems
WO1999032970A1 (en) * 1997-12-23 1999-07-01 Ab Initio Software Corporation Method for analyzing capacity of parallel processing systems

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1998052122A1 (en) * 1997-05-14 1998-11-19 Compuware Corporation Accurate profile and timing information for multitasking systems
WO1999032970A1 (en) * 1997-12-23 1999-07-01 Ab Initio Software Corporation Method for analyzing capacity of parallel processing systems

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
IBM Technical Disclosure Bulletin "Optimization of computer configurations"published 1975, Vol 18 No. 7, Angiulli et al, pages 2235 - 2237 *
Varki E. and Dowdy L., "Proceedings of MASCOT 96'",1996, pages 291-295, "Response time analysis of two server fork-join systems" *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015011487A1 (en) 2013-07-24 2015-01-29 David Mchattie Monitoring the performance of a computer

Also Published As

Publication number Publication date
GB0120564D0 (en) 2001-10-17
JP2002123403A (en) 2002-04-26

Similar Documents

Publication Publication Date Title
US7054943B1 (en) Method and apparatus for dynamically adjusting resources assigned to plurality of customers, for meeting service level agreements (slas) with minimal resources, and allowing common pools of resources to be used across plural customers on a demand basis
US6901484B2 (en) Storage-assisted quality of service (QoS)
US20190306255A1 (en) System and method for throttling service requests having non-uniform workloads
US20030005028A1 (en) Method and apparatus for controlling the number of servers in a hierarchical resource environment
US8224465B2 (en) Coordinating service performance and application placement management
USRE42726E1 (en) Dynamically modifying the resources of a virtual server
JP5041805B2 (en) Service quality controller and service quality method for data storage system
US8635625B2 (en) Power-aware workload allocation in performance-managed computing environments
US9178763B2 (en) Weight-based collocation management
US20050060456A1 (en) Method and apparatus for multi-port memory controller
US7032222B1 (en) Method and system for determining resource allocation to users by granting request based on user associated different limits and resource limit
US20150156123A1 (en) System and method for throttling service requests using work-based tokens
US20080162965A1 (en) Managing performance of a processor in a data processing image
JP2006511137A (en) Flow control in network devices
US10069757B1 (en) Reserved network device capacity
JP4530285B2 (en) Network accommodation design system, method for determining concentration ratio thereof, program, and storage medium therefor
GB2370140A (en) A delay accounting method for computer system response time improvement
CN109144664B (en) Dynamic migration method of virtual machine based on user service quality demand difference
CN109586971B (en) Load resource demand evaluation method based on linear relation
Kang et al. Star: secure real-time transaction processing with timeliness guarantees
CA2318082C (en) Method and device for controlling processes on a computer system
CN106445661A (en) Dynamic optimization method and system
CN114489463A (en) Method and device for dynamically adjusting QOS (quality of service) of storage volume and computing equipment
Mahabhashyam et al. Dynamic resource allocation of shared data centers supporting multiclass requests
CN116582497B (en) Method for shaping self-adaptive traffic of high-efficiency GIS service under single server condition

Legal Events

Date Code Title Description
WAP Application withdrawn, taken to be withdrawn or refused ** after publication under section 16(1)