System and method for emulating input/output performance of an application
View Patent ↗A system and method for emulating the input/output performance of an application. A workload description language is used to produce a small but accurate model of the application, which is flexible enough to emulate the application's performance with varying underlying system configurations or operating parameters. The model describes I/O operations performed by the application, and reflects any dependencies that exist between different application threads or processes. The model is then executed or interpreted with a particular system configuration, and various parameters of the I/O operations may be set at the model's run-time. During execution, the input/output operations described in the model are generated according to the specified parameters, and are performed. The system configuration and/or I/O operation parameters may be altered and the model may be re-run.
1. A computer-implemented method of emulating an input/output performance of an application, comprising:
obtaining a model of the application's workload, wherein said model describes the application's input/output behavior within different layers of an input/output stack, the input/output behavior comprising:
input/output operations performed by the application; and
input/output dependencies between application threads or processes, wherein an input/output dependency occurs when an input or output operation from one thread or process blocks an input or output operation from another thread or process;
setting one or more parameters for said input/output operations;
setting one or more parameters in the model for a computing environment
under which the model is executed or interpreted; and interpreting said model to:
generate said input/output operations configured according to said parameters; and
perform said generated input/output operations.
2. The method of claim 1 , wherein said dependency involves one of the two or more threads blocking on another of the two or more threads.
3. The method of claim 1 , further comprising, prior to said obtaining.
executing the application;
tracing input/output performed by the application during the execution; and
creating said model to describe said traced input/output.
4. The method of claim 3 , wherein said tracing comprises:
capturing all input/output operations performed by the application during said execution;
identifying characteristics of the application, said characteristics including at least one of the following:
number of application processes;
number of executing threads;
number of files manipulated;
types of input/output operations performed; and
quantity of each type of input/output operation performed.
5. The method of claim 3 , wherein said creating comprises:
identifying types of input/output operations performed during said execution; and
for each type of input/output operation, statistically reducing the identified input/output operations of said type into a probability distribution.
6. The method of claim 3 , further comprising:
during said executing of the application, capturing summary statistics describing the traced input/output; and
after said interpreting:
capturing emulation statistics describing said performed input/output operations; and
comparing said summary statistics with said emulation statistics.
7. The method of claim 1 , further comprising:
changing one or more of said parameters; and
re-interpreting said model to generate said input/output operations having the changed parameters.
8. The method of claim 1 , further comprising:
changing a filesystem on which said input/output operations are performed.
9. The method of claim 1 , further comprising:
changing an operating system on which said input/output operations are performed.
10. The method of claim 1 , further comprising:
changing an access method by which said input/output operations are performed.
11. The method of claim 1 , wherein said model comprises:
one or more model processes, each said model process having one or more process parameters and comprising a set of model threads;
wherein each said model thread has one or more thread parameters and comprises a series of flow operations, wherein each of the flow operations is a predefined operation; and
wherein each said flow operation has one or more operation parameters and describes one of said input/output operations.
12. The method of claim 1 , wherein the one or more parameters for the computing environment comprise at least one of:
a file system,
an operating system,
a storage device, or
one or more communication interfaces.
13. A non-transitory computer readable medium storing instructions that, when executed by a computer, cause the computer to perform a method of emulating an input/output performance of an application, comprising:
obtaining a model of the application's workload, wherein said model describes the application's input/output behavior within different layers of an input/output stack, the input/output behavior comprising:
input/output operations performed by the application; and
input/output dependencies between application threads or processes, wherein an input/output dependency occurs when an input or output operation from one thread or process blocks an input or output operation from another thread or process;
setting one or more parameters for the input/output operations;
setting one or more parameters in the model for a computing environment under which the model is executed or interpreted; and
interpreting said model to:
generate said input/output operations configured according to said parameters; and
perform said generated input/output operations.
14. The computer-readable medium of claim 13 , wherein the one or more parameters for the computing environment comprise at least one of:
a file system,
an operating system,
a storage device, or
one or more communication interfaces.
15. A computer system for modeling an application's input/output, comprising:
a processor;
a model describing the application's input/output behavior within different layers of the input/output stack, the input/output behavior comprising:
input/output operations performed by the application, and
input/output dependencies between application threads or processes, wherein an input/output dependency occurs when an input or output operation from one thread or process blocks an input or output operation from another thread or process;
a set of input/output operation parameters;
a set of computing environment parameters under which the model is executed or interpreted;
wherein the processor is configured as an interpreter that generates the input/output operations configured according to the set of parameters; and
a file system configured to perform said input/output operations on a set of files described by the mode.
16. The computer system of claim 15 , wherein:
said interpreter generates the input/output operations in a manner reflecting said dependencies.
17. The computer system of claim 15 , further comprising:
a set of operation files, wherein said interpreter invokes one of the operation files for each type of input/output operation generated.
18. The computer system of claim 15 , further comprising:
a collection of storage devices storing the set of files.
19. The computer system of claim 15 , wherein the one or more parameters for the computing environment comprise at least one of:
a file system,
an operating system,
a storage device, or
one or more communication interfaces.