I need the list of parameters to be checked for the performance issue in the FSM Application.
Solved
What ae the parameters need to be checking in FSM for performance issue
Best answer by ChristineLaVoi
| Unfortunately, to resolve you are going to first define, performance, below is the questions support will ask: | |
| 1 | Get an architectural diagram of the customer's environment. This includes: |
| 1a | Application Servers |
| 1b | Web Servers |
| 1c | Database Servers |
| 1d | Miscellaneous items such as firewalls, routers, etc. |
| 1e | Identify which servers are physical vs. virtual. For virtual servers, identify what physical servers they are running on. |
| 1f | Identify what other software is running on the servers. |
| 1g | Identify the key hardware characteristics of the servers (memory, processors, storage). |
| 1h | Identify the operating system versions. |
| 1i | Identify the database type and version. |
| 2 | What does the customer mean by "bad performance"? |
| 2a | Is performance universally bad or bad only at certain times? |
| 2b | Are they receiving time-outs from the smart client? |
| 2c | Are screens taking a long time to launch? |
| 2d | Does it take longer to select data from the system? |
| 2e | Does it take longer to update data in the system? |
| 2f | Is poor performance limited to certain screens or processes? |
| 2g | Are only certain types of users experiencing the problems? |
| 3 | If performance varies, under what circumstances does it suffer? This includes such things as: |
| 3a | Does it degrade at certain times of the day consistantly? If so, what is happening at that time? |
| 3ai | Integrations |
| 3aii | Reports |
| 3aiii | Increased users on the system |
| 3aiv | Intensive processes (picking, etc) |
| 3b | Does it degrade in response to users doing something specifically? You'll probably need to track the activities happening in the system across multiple occurrences of the performance issues to detect a trend. Users may do the same thing at different times on different days. |
| 3bi | Integrations |
| 3bii | Reports |
| 3biii | Intensive processes (picking, etc) |
| 4 | When performance degrades, what does hardware utilization on the servers look like? (use tools such as Windows Task Manager on the App, Web and Database Servers and database specific tools such as SQL Profiler and Actvitiy Monitor on the Database Server) |
| 4a | Memory usage (periodic spikes are generally ok, but sustained high levels of usage or growth typically indicate a problem) |
| 4b | CPU usage (periodic spikes are generally ok, but sustained high levels of usage typically indicate a problem) |
| 4c | Existance of database locks or excessively long running queries. Deadlocks are always a problem, but even high volumes of other locks can lead to a slowdown. For queries, any query running for over 10 seconds is a definite problem, and any query consistantly running for over 1 second maybe a problem. This also needs to be weighed against the # of occurrences (a query taking 5 seconds occurring 10,000 times a day is more significant than a query taking 30 seconds running 3 times a day). |
| 4d | There are other potential issues such as firewalls, routers, network bandwidth, etc. that should be considered as well. |
+24Hi
Have you tried a search of the Community and Knowledge articles? There have been a number of posts and articles written regarding FSM performance an how to track down and improve this.
If you identify something not covered by the existing documentation please let us know and we can look to fill that gap.
Kind regards,
Lee Pinchbeck
Currently reading: The Way of Kings - Brandon Sanderson
1 person likes this
| Unfortunately, to resolve you are going to first define, performance, below is the questions support will ask: | |
| 1 | Get an architectural diagram of the customer's environment. This includes: |
| 1a | Application Servers |
| 1b | Web Servers |
| 1c | Database Servers |
| 1d | Miscellaneous items such as firewalls, routers, etc. |
| 1e | Identify which servers are physical vs. virtual. For virtual servers, identify what physical servers they are running on. |
| 1f | Identify what other software is running on the servers. |
| 1g | Identify the key hardware characteristics of the servers (memory, processors, storage). |
| 1h | Identify the operating system versions. |
| 1i | Identify the database type and version. |
| 2 | What does the customer mean by "bad performance"? |
| 2a | Is performance universally bad or bad only at certain times? |
| 2b | Are they receiving time-outs from the smart client? |
| 2c | Are screens taking a long time to launch? |
| 2d | Does it take longer to select data from the system? |
| 2e | Does it take longer to update data in the system? |
| 2f | Is poor performance limited to certain screens or processes? |
| 2g | Are only certain types of users experiencing the problems? |
| 3 | If performance varies, under what circumstances does it suffer? This includes such things as: |
| 3a | Does it degrade at certain times of the day consistantly? If so, what is happening at that time? |
| 3ai | Integrations |
| 3aii | Reports |
| 3aiii | Increased users on the system |
| 3aiv | Intensive processes (picking, etc) |
| 3b | Does it degrade in response to users doing something specifically? You'll probably need to track the activities happening in the system across multiple occurrences of the performance issues to detect a trend. Users may do the same thing at different times on different days. |
| 3bi | Integrations |
| 3bii | Reports |
| 3biii | Intensive processes (picking, etc) |
| 4 | When performance degrades, what does hardware utilization on the servers look like? (use tools such as Windows Task Manager on the App, Web and Database Servers and database specific tools such as SQL Profiler and Actvitiy Monitor on the Database Server) |
| 4a | Memory usage (periodic spikes are generally ok, but sustained high levels of usage or growth typically indicate a problem) |
| 4b | CPU usage (periodic spikes are generally ok, but sustained high levels of usage typically indicate a problem) |
| 4c | Existance of database locks or excessively long running queries. Deadlocks are always a problem, but even high volumes of other locks can lead to a slowdown. For queries, any query running for over 10 seconds is a definite problem, and any query consistantly running for over 1 second maybe a problem. This also needs to be weighed against the # of occurrences (a query taking 5 seconds occurring 10,000 times a day is more significant than a query taking 30 seconds running 3 times a day). |
| 4d | There are other potential issues such as firewalls, routers, network bandwidth, etc. that should be considered as well. |
I found the following link for the performance tuning
Kindly share there are any other documents related to performance tuning.
+13
Also you might find very useful these document /recordings in \\mkenas21\metrix$\RnD\Public\Performance
Ann Degroat wrote:
Also you might find very useful these document /recordings in \\mkenas21\metrix$\RnD\Public\Performance
Can we have access to these files?
+13Below are the documents that may be useful.
Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.