on 05-24-2012 3:13 PM
Hi there.
Since three weeks we get a red EWA alert for the Java System Data in our Solution Manager.I ínstalled the SMD Diagnostics, incl. SMD Agent standalone 7.30 on top,Wily introscope Enterprise manager and Agent 8.24. SLD was activated and installed in this case also.
The parameters for J2ee Engine was related to the SAP Notes checked:
SAP Note 723909 - Java VM settings for J2EE 6.40-7.0
sapnote_861215 Empfohlene Einstellungen für Linux
in AMD64EM64T JVM
sapnote_1008311 Empfohlene Einstellungen für NW 7.0=
SR2 für AIX JVM (J9)
sapnote_1149214 Wily Introscope Agent mit IBM Java
VM 1.4.2 (SR1x) x86_64
sapnote_1303953 SAP JVM memory parameters and their names in the Config
Tool
and Hinweis 1509719 -
LM-SERVICE 701 SP08 Patch 1 so on and so on.
Patched LM service etc. deployed.
Was also checked and the parameters adapted.
All is not the solution for our system. I
installed 50 % more RAM, increased the heap as recommended from 2048 MB To 3048
for the JVM , get a RED EWA ALERT. Checked all JAVA parameters again, checked std_server0.out.
WHATS NOW??? The EWA shows for Garbage Collection Time hourly data
the green alert, for fraction of Garbage Collections 100 %????
So, we are near to be finished, only got to install on the managed
systems the agents and connect to SMD Diag. Now we have the RED EWA Alert on
our Solution manager 7.01 EHP 1 with LINUX, MAXDB 7.8 and IBM J9 VM (build 2.3,
J2RE 1.4.2 IBM J9 2.3 Linux amd64-64 j9vmxa64142ifx-20110312 (JIT enabled)
Update to SAP JVM later, related to 7.20 Kernel and SLES11. I adapted the
parameters last weeks and get also a red EWA alert for the garbage collection
again. The third.
So, how I could go on and solve that error of EWA?? I am near to think it’s
a EWA problem. Thx Ralf
SNOTE 1487678 - EarlyWatch Alert - Issues in ST-SER 701_*
is not rerally the solution for "Fraction of Full Garbage Collections"
Part 1 describes the error,
2. You are getting Red SAP EarlyWatch Alert for systems with a Java stack due to
Garbage collection time (which is showing as Red). In the check "Garbage
Collection Time" the max values are above 100% which is obviously wrong.
The
values of "%GC Time (Last 5 Minutes)" available on the EM server or in the RCA
infocube are not correct. Sometimes the value exceeds 100%.
Error in the
calculation of "% GC Time last 5 minutes", that is collected by the Wily IS
hostagent. According to the documentation of the Wily IS EM this KPI means the
following: "The chart shows in percentage how much time of the last one or five
minutes was spent by the java virtual machine (JVM) for garbage collection (GC).
For a well behaved JVM this value should not increase permanently and should
remain below 5%."
The wrong calculation(or values above 100%) leading to the
false conclusion there is a problem with Garbage Collection. Manual inspection
of the std_server0 out file showed there was no problem with GC.
Reason and Prerequisites
2. For Sun and HP JVM, missing VM parameter : -XX:+PrintGCTimeStamps will cause
that GC log is not correctly parsed. Workaround is to add the parameter in
Server VM parameters.
Done before!!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
80 | |
24 | |
12 | |
9 | |
7 | |
6 | |
5 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.