Performance testing

For Performance meetup at SolarWinds

Transcript

1. @jakubfojtl Performance TESTS @ysoftdevs
2. Agenda
3. Agenda Motivation
4. Agenda Motivation The beginnings in Y Soft
5. Agenda Motivation The beginnings in Y Soft Nowaday’s stack
6. Agenda Motivation The beginnings in Y Soft Nowaday’s stack What’s next
7. Motivation
8. Our motivation
9. On-site deployments Our motivation
10. Sizing of our software On-site deployments Our motivation
11. Sizing of our software Aka How many servers for my users On-site deployments Our motivation
12. Cloud?
13. Cloud? It depends…
14. Cloud? It depends…
15. Cloud? It depends… or
16. Cloud? It depends… or well-operated = predictable
17. Beginnings
18. SafeQ
19. SafeQ
20. SafeQ CTRL+P
21. SafeQ CTRL+P
22. SafeQ CTRL+P
23. SafeQ CTRL+P
24. SafeQ CTRL+P
25. SafeQ CTRL+P Secure print
26. Installation
 and
 evaluation
27. Installation
 and
 evaluation
28. Installation
 and
 evaluation Order VMs from internal tesbed
29. Installation
 and
 evaluation Install and configure SafeQ using installers Order VMs from internal tesbed
30. Installation
 and
 evaluation Install and configure SafeQ using installers Order VMs from internal tesbed Do the magic*
31. Installation
 and
 evaluation Install and configure SafeQ using installers Order VMs from internal tesbed Do the magic* Spend weeks gathering data and evaluate
32. Installation
 and
 evaluation Install and configure SafeQ using installers Order VMs from internal tesbed Do the magic* Spend weeks gathering data and evaluate Ok, X printers per server
33. *Do the magic
34. *Do the magic
35. *Do the magic Many printing protocols (LPR, IPP…)
36. *Do the magic Many printing protocols (LPR, IPP…) Our own HW
37. *Do the magic Many printing protocols (LPR, IPP…) Our own HW Many Vendor SDKs
38. *Do the magic Many printing protocols (LPR, IPP…) Our own HW Many Vendor SDKs In-house SW to the rescue
39. *Do the magic Many printing protocols (LPR, IPP…) Our own HW Many Vendor SDKs In-house SW to the rescue
40. *Do the magic Many printing protocols (LPR, IPP…) Our own HW Many Vendor SDKs In-house SW to the rescue Fufator
41. New Requirements
42. New Requirements
43. New Requirements Able to describe customer’s environments
44. New Requirements Able to describe customer’s environments Monitoring
45. New Requirements Able to describe customer’s environments Reusable at customer Monitoring
46. New Requirements Able to describe customer’s environments Reusable at customer Automatic from start to end Monitoring
47. New Requirements Able to describe customer’s environments Reusable at customer Automatic from start to end Monitoring Flexible in test scenarios
48. Today
49. Stack
50. Infrastructure Terraform Packer Stack
51. Infrastructure Terraform Packer Application deployment Chef Stack
52. Infrastructure Terraform Packer Application deployment Chef Load generation Locust Stack
53. Infrastructure Terraform Packer Application deployment Chef Load generation Locust Monitoring Grafana PowerBI Stack
54. Architecture
55. Terraform
56. Terraform“vagrant for your datacenter”
57. Terraform“vagrant for your datacenter” yaml-based configuration
58. Terraform“vagrant for your datacenter” yaml-based configuration every resource of underlying
 provider is a module
59. Terraform“vagrant for your datacenter” yaml-based configuration every resource of underlying
 provider is a module describes anything from disk to network
60. Chef
61. Chef ruby-based DSL for deployments
62. Chef ruby-based DSL for deployments Artifactory for artefacts
63. Chef ruby-based DSL for deployments Artifactory for artefacts first class support for DSC
64. Chef ruby-based DSL for deployments Artifactory for artefacts first class support for DSC Chef server <- chef client
65. Chef ruby-based DSL for deployments Artifactory for artefacts first class support for DSC Chef server <- chef client we plan to reuse in installers
66. Locust
67. Locust yaml-based configuration
68. Locust yaml-based configuration Load scenarios
69. Locust yaml-based configuration Load scenarios Server-agent architecture
70. Evaluation
71. Evaluation defined scenario finished
 within constraints
72. Evaluation defined scenario finished
 within constraints e.g. printing <job> for 90th percentile of <users> finishes in <200 ms>
73. Evaluation defined scenario finished
 within constraints e.g. printing <job> for 90th percentile of <users> finishes in <200 ms> HW utilization is secondary
76. What’s next
77. What’s next
78. What’s next reuse this at customers
79. What’s next reuse this at customers adding more user focused metrics
80. What’s next reuse this at customers adding more user focused metrics use chef to provision 3rd party systems
81. What’s next reuse this at customers adding more user focused metrics use chef to provision 3rd party systems overview in PowerBI
82. @jakubfojtl | jfojtl.com Thank you