To be straight forward, i do not see any problem in posting the key tech notes for the bloggers/people who are in search of information on Google/Insight into it day to day basis. But quite a few months back, i know that the XXXX Forum administrator came as dummy user id and was scolding me 🙂 ~ as i was pointing him/her/them for their incorrect info that they show in forum or couple other such activities where forum patrol agents need to mark posts as recommended (which was not happening for years and i initiated/requested them to follow/etc).
If there are such people you come across in life, i would simply say ignore and do your best always. I was not arrogant or so in forum ~ hope so (If i was, please excuse for the same). I made or advised them to keep the forum effective so that, it will help other users who uses it. I told the forum admin that, wouldn’t post any post after that and i was following my words till date. But i feel now that, why do i need to stop my thirst of learning new things daily for someone who speaks like it. The story was quite a long one 🙂 thanks for reading it.
——————————————————————————————–
Let’s come to the point of troubleshooting the MicroStrategy web crash issue.
From my experience,
==> I’ll be working on the below points before i look back into the MicroStrategy Knowledge base.
1) Identify whether Web Server is up and running
2) Goto Web Admin page and check the logs in web to determine any insight about problem (like due to specific report execution/memory/space/plugins changes/weird behavior if any etc..)
3) Goto DSSErrors.log and verify if there are info available to troubleshoot.
4) Take necessary action from the logs analyzed.
But, when your an MicroStrategy Architect / Principal ~ You should think beyond that as well. Which means, you need to know about the
* Version of MicroStrategy software
* OS in which it is installed
* JVM parameters set
* JAVA version
* Concurrency of users accessing the application during that time
* Job Prioritization if available or not
* Logs enabled or not
* Health Centre setup or not to have a look
Lot more to analyze in depth when you dig or work on an issue. This will be known when you have experience/hands on to the core of the servers in and out.
—————————————————————————————————-
Below was the information that MicroStrategy suggests in the Knowledge Base site.
- Fill in the following table
|
FEATURE
|
VALUE
|
|
|
APP SERVER MACHINE
|
Application Server Version
|
|
|
Operating System Version
|
||
|
Operating System – 32bit or 64bit ?
|
||
|
MicroStrategy Web Version
|
||
|
Total Physical memory
|
||
|
Processor type
|
||
|
Number of processors
|
||
|
Java Version
|
||
|
Java 32bit or 64bit?
|
||
|
Current Java min heap size
|
||
|
Current Java max heap size
|
||
|
Number of threads configured on the application server
|
||
|
Other Java Parameters set
|
||
|
Number of JVM instances on the machine
|
||
|
Memory available on the machine after starting the application server
|
||
|
Memory available on the machine during runtime/peak times
|
||
|
Average concurrency on peak times
|
||
|
AVG CPU Usage during peak times
|
||
|
Definition of the objects involved (if applicable)
|
- Five thread dumps: for every 30 seconds , take a thread dump ( do this four times) and then take another thread dump after 2 minutes ( total of five thread dump) .
- Collect Heapdump when hanging or crash issue occurs
-
Refer TN: 35733 : How to generate a Memory Heap dump.
-
- Collect Verbose GC logs.
-
Refer TN: 10887 : How to turn on the Verbose Garbage collection logging for the Java Virtual Machine
-
Note: Enable Verbose GC and heapdump prior to issue occurs.
-
- Collect the web application log from application server. (For example: catalina.out for Apache Tomcat)