You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/mysql/single-server/concept-perf-benchmark-best-practices.md
+5-5Lines changed: 5 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -190,8 +190,8 @@ If you can’t easily deploy a copy of your application in Azure, you'll need to
190
190
191
191
There are two main approaches to generating the test payload (SQL statement samples):
192
192
193
-
1. Observe/record the SQL traffic occurring in your current database, then generate SQL samples based on those observations. For details on how to record query traffic by leveraging a combination of audit logs and slow query logging in Azure Database for MySQL.
194
-
2. Use actual query logs as the payload. Third party tools such as “Percona Playback” can generate multi-threaded workloads based on MySQL Slow Query Logs.
193
+
- Observe/record the SQL traffic occurring in your current database, then generate SQL samples based on those observations. For details on how to record query traffic by leveraging a combination of audit logs and slow query logging in Azure Database for MySQL.
194
+
- Use actual query logs as the payload. Third party tools such as “Percona Playback” can generate multi-threaded workloads based on MySQL Slow Query Logs.
195
195
196
196
If you decide to generate SQL sample manually, be sure that the sample contains:
197
197
@@ -226,10 +226,10 @@ Assume that you decide to run benchmarks against a db.r3.2xlarge server and find
226
226
227
227
### Measure both sustained and peak performance
228
228
229
-
The test strategy you choose should provide you with answers to the following two questions:
229
+
The test strategy you choose should provide you with answers to whether the database will provide adequate:
230
230
231
-
1. Will the database provide adequate sustained performance? In other words, will it perform as expected under the normal workload, when user traffic is smooth and well within expected levels?
232
-
2. Will the database provide adequate peak performance? That is, will it ensure application responsiveness during traffic spikes?
231
+
- Sustained performance - Will it perform as expected under the normal workload, when user traffic is smooth and well within expected levels?
232
+
- Peak performance - Will it ensure application responsiveness during traffic spikes?
0 commit comments