Capping dev_ms trace file size

hi - i'm wondering if anyone knows a way to help. we have been asked by SAP to run our msg server at an elevated trace level (trace level 3 - we set it from SMMS). this writes out a huge amt of data to the dev_ms trace file.
the default trace file (dev*) size, per the rdisp/TRACE_LOGGING param default is "on, 10m" (i.e. 10mb).
that applies to "all" the dev* trace files obviously.
w/ our elevated dev_ms trace set, it's overwrapping in ~3 min ...so between dev_ms.old and dev_ms, we never have more than ~10 min of logs kept at any one time.
the pt of us running at this elevated dev_ms trace level is so we can capture (save off) the trace file and send to SAP the next time our msg server crashes.
our SAP file system mountpoint /usr/sap/<SID> is limited in size...and by setting rdisp/TRACE_LOGGING value to a higher value it affects all the dev* files, not just the 1 file i really care about increasing the capped file size on (dev_ms).
**QUESTION*: does any one know a way i could keep dev_ms capped at a large value like 100mb yet keep all the other dev files at the normal 10mb default? thanks in advance

1. Increase rdisp/TRACE_LOGGING to 100MB.
2. Set (SM51) > Select All Processes > Menu > Process > Trace > Active Components > Uncheck everything and set trace level to 1.
3. Menu > Process > Trace > Dispatcher > Change Trace Level > Set to 2
Wouldn't this essentially just increase dev_ms to 100MB while leaving other dev* trace files to not log anything?

Similar Messages

Trace file size reached the max size

Hi All,
This morning we had a issue with the oracle trace file. When I looked the alert log, i figured that there was an error saying Trace file reached max size. Question is what happens to the processes that tries to write to the trace file when the trace file was full?
My thinking is, the process, that tries to write to trace file, will hang until the trace file size is incresed. Am I correct. Please share you thoughts
Thanks
--

Process will continue, it will not hang. Trace information will not be written once the trace file limit is reached.
HTH
Srini

Trace file size is huge

HI Experts,
In my production database one trace file alone getting created in huge size, could you please make me aware of the reason behind this and solution to this.
Path:/oracle/admin/proddb/bdump
proddb_j000_17828.trc /5G/
Regards,
Salai

You can limit trace file size using MAX_DUMP_FILE_SIZE db parameter if you have a free space problem too.
http://download.oracle.com/docs/cd/B19306_01/server.102/b14237/initparams116.htm#REFRN10103
But of course this wont solve your original issue. Try to analyze content of the trace file and try to solve the base issue.

Trace file size

Hi,
I have ECc 6.0 System with DB2 Database
I have only ABAP instance. Please let me know of how to view/change the maximum SQL trace file size generated from ST05
I checked the forums and it refers to the following link, which says i should have JAVA
[http://help.sap.com/saphelp_nw04/helpdata/en/1d/014da5fb814e0299ef752b5696979f/content.htm]
Please help

Hello,
For ABAP you need to use the following parameters :
[http://help.sap.com/saphelp_nw04/helpdata/en/1f/8311594bc511d189750000e8322d00/content.htm|http://help.sap.com/saphelp_nw04/helpdata/en/1f/8311594bc511d189750000e8322d00/content.htm]
Success.
Wim

Any way to limit DV captured file size?

Hello all -
When capturing from a DV camcorder when the source is not timecoded, is there a way to cap the max file size? Ideally I'd like to set a size limit, eg. 650mb and then have the files split into project 01.avi, project 02.avi, etc.
Thanks,
Steve

So I gather the answer is No, Premiere Elements cannot split AVI files when capturing from DV camcorders, except by timecode?
I realize there are other packages I can pay for that support the feature but I expect Premiere Elements (at least by v8) to have that capability.
In any event, Easy Video Splitter (doeasier.org) is now free and does the job well.

Deleting default trace file in java

Hello all,
Whats the best way to automatically delete default trace file in java instance? I want to delete these files lets say every week or 2 wkkds,Is there any type of configuration that can be done through NWA or OS...? or do i have to write some scripts? Please advice.

Perfect.
When set to 10, 0 -->9 are cyclic files. Regarding archiving, yes this is possible. Apart from defaulttrace files you can also archive other java log files (system, network,security etc). The configuration procedure is explained :-
http://help.sap.com/saphelp_nw04/helpdata/en/48/2edfd5bd3e0d4a81b90325fe195a70/frameset.htm
Once the 10th trace file(i.e.defaultrace9) is full, all the trace files (from defaulttrace0-defaultrace9) are zipped to a single file. The file compression ratio is extremely good.
Unfortunately there is no automated mechanism to clear off the archived files. This needs manual intervention(regular cleanup, script etc).
The archived log files may also be viewed for analysis from NWA.
While estimating the size for each defaulttrace consider three factors i) System load ii) Number of defaulttrace files.
iii) Retention days(without archiving)
Ex: System load as observed :
Say the trace file size is 10MB and 6 trace files get written in 4 hours of peak time and 1 trace files is written for 4 hours during non-peak time .Assuming 8 hours of peak business hours and rest all non-working hours.
No of files generated is 62 + 14 = 16 files of 10 MB each are generated per week day
If the defaulttrace file count is set to 16, you can retain one days' logs . Say you wish to retain the same with less number of files (i.e. 8) you would have to increase the file size to 20 MB. Also will have to look at file system space availability.
CAUTION: For a file system occupancy estimate, you will have to also consider as to how many Server nodes are
configured as each node has its own log area !
For this you need to observe the pattern as to how many files are being written then estimate the size and count
cheers !
PRADi

Trace file of directory /ORACLE/ SID /SAPTRACE/USERTRACE increase ceaseless

Dear all,
Now I face a problem that the Trace file of directory /ORACLE/<SID>/SAPTRACE/USERTRACE increase ceaseless and very quickly, the trace files size was increased to 8G only two days, so the directory always is full due to this.
So anybody could tell me this is why ? and what method could help to decrease the trace log to produce ?
Thanks & Regards,
Michael

>
mho wrote:
> There could be various issues causing this. I recommend having a look at [1505012 - Unrequired Oracle trace data in R3trans and tp|https://service.sap.com/sap/support/notes/1505012], for an bug in R3trans and tp.
>
> If this does not match, then please tell us what's inside the files.
>
> Cheers Michael
Thanks for your reply, I saw the notes and use Transaction SM50 to check the trace level, the level is default ,
this system is our solution manager, although the directory is full but the system still could be connected.
The trace files contents as below：
u201C*** 2011-04-25 12:22:41.995
ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [qertbFetchByRowID], [], [], [], [], [], [], []
Current SQL statement for this session:
SELECT * FROM "TBTCO" WHERE "JOBNAME" = :A0 AND "JOBCOUNT" = :A1
Call Stack Trace -
calling              call     entry                argument values in hex
location             type     point                (? means dubious value)
_ksedst38           CALLrel _ksedst10           0 1
_ksedmp898          CALLrel _ksedst0            0
_ksfdmp14           CALLrel _ksedmp0            3
_kgerinv+140         CALLreg 00000000             32560400 3
_kgeasnmierr19      CALLrel _kgerinv0           32560400 9548210 38FFDE0 0
                                                   ECDD670
__VInfreq__qertbFet CALLrel _kgeasnmierr+0       32560400 9548210 38FFDE0 0
chByRowID+2583
_opifch2+3115        CALL??? 00000000             2AFAEC3C 1E9B2F4 ECDD9D8 2
_opiefn0348         CALLrel _opifch20           89 4 ECDDB7C
_opiefn21           CALLrel _opiefn00           4E 4 ECDF698 0 0 0 0 0 0 0
_opiodr+1099         CALLreg 00000000             4E 4 ECDF698
_ttcpip+1273         CALLreg 00000000             4E 4 ECDF698 C
_opitsk+1017         CALL??? 00000000
_opiino1087         CALLrel _opitsk0            0 0
_opiodr+1099         CALLreg 00000000             3C 4 ECDFC30
_opidrv819          CALLrel _opiodr0            3C 4 ECDFC30 0
_sou2o45            CALLrel _opidrv0            3C 4 ECDFC30
opimaireal112     CALLrel _sou2o0             ECDFC24 3C 4 ECDFC30
_opimai92           CALLrel opimaireal0       2 ECDFC5C
_OracleThreadStart@ CALLrel _opimai+0
4+708
77E66060             CALLreg 00000000
Binary Stack Dump -
========== FRAME [1] (_ksedst38 -> _ksedst10) ==========
Dump of memory from 0x0ECDD544 to 0x0ECDD554
ECDD540          0ECDD554 0040467B 00000000      [T...{F@.....]
ECDD550 00000001                             [....]
========== FRAME [2] (_ksedmp898 -> _ksedst0) ==========
Dump of memory from 0x0ECDD554 to 0x0ECDD614
Could you help to check this issue ?
Thanks

Generation of MMON process trace files in large file size (GB Size)

Hi,
I have created a database using the dbca in windows platform. Few days I found that, in the BDUMP directory the MMON process trace files are getting generated. The files starts to generate in MB size and will increase upto GB size. I know that the back ground process trace files cannot be disabled. So now iam force to manually delete these files from the bdump directory.plz help me to resolve this issue.
I have checked and verified the SGA size, Shared Pool size and other memory areas.
The statistics level in Typical also.
But still the files are generated.
PLease Helppp.....
Shiyas

hi
As per your instruction i have checked the Alert log file. I have pasted a part of errors that found in the alert log file.
Mon Jun 07 09:30:58 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_652.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:31:02 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_652.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:36:00 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_652.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:36:08 2010
Restarting dead background process MMON
MMON started with pid=11, OS id=656
Mon Jun 07 09:36:11 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:36:15 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:41:12 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:41:16 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:46:13 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:46:17 2010
Errors in file d:\oracle\product\10.2.0\admin\mir\bdump\mir_mmon_656.trc:
ORA-00600: internal error code, arguments: [kjhn_post_ha_alert0-862], [], [], [], [], [], [], []
Mon Jun 07 09:50:18 2010
Shutting down instance: further logons disabled
Mon Jun 07 09:50:19 2010
Stopping background process QMNC
Mon Jun 07 09:50:19 2010
Stopping background process CJQ0
Mon Jun 07 09:50:20 2010
Stopping background process MMNL
Mon Jun 07 09:50:21 2010
Stopping background process MMON
Mon Jun 07 09:50:22 2010
Shutting down instance (immediate)
License high water mark = 4
Mon Jun 07 09:50:22 2010
Stopping Job queue slave processes, flags = 7
Mon Jun 07 09:50:22 2010
Job queue slave processes stopped
Waiting for dispatcher 'D000' to shutdown
All dispatchers and shared servers shutdown
Mon Jun 07 09:50:23 2010
alter database close normal
Mon Jun 07 09:50:23 2010
SMON: disabling tx recovery
SMON: disabling cache recovery
Mon Jun 07 09:50:23 2010
Shutting down archive processes
Archiving is disabled
Archive process shutdown avoided: 0 active
Thread 1 closed at log sequence 71
Successful close of redo thread 1
Mon Jun 07 09:50:23 2010
Completed: alter database close normal
Mon Jun 07 09:50:23 2010
alter database dismount
Completed: alter database dismount
ARCH: Archival disabled due to shutdown: 1089
Shutting down archive processes
Archiving is disabled
Archive process shutdown avoided: 0 active
ARCH: Archival disabled due to shutdown: 1089
Shutting down archive processes
Archiving is disabled
Archive process shutdown avoided: 0 active
But I am not able to understand anything above of this.
And I am sorry we dont have the metalink support or srs support.
Is there any other way to resolve this issue.
Shiyas

Best practice to reduce size of BIA trace files

Hi,
I saw alert on BIA monitor says 'check size of trace files'. Most of my trace files are above 20MB. I clicked on details it says "Check the size of your trace files. Remove or move the trace files with the memory usage that is too high or trace files that are no longer needed."
I would like to reduce them these trace files but not sure what is the safetest way to do it. Any suggestion would be appreciated!
Thanks.
Mimosa

Mimosa,
Let's be clear here first. The tracing set via sm50 is for tracing on the ABAP side of BI not the BIA.
Yes, it is safe to move/delete TrexAlertServer.trc, TrexIndexServer.trc, etc from the OS level. You can also right click the individual trace when you enter the "Trace" tab in the TREX Admin Tool (python) and I believe there is options to delete them there but it is certaintly OKAY to do this on the OS level. They are simply recreated when new traces are generated.
I would recommend that you simply .zip the files and move the .zip files to another folder in case SAP support may need them to analyze an issue. As long as they aren't huge, and if hard disk space permits, this shouldnt be an issue. After this you then will need to delete the trace file. It is important that if a trace file has an open handle registered to it then it wont let you delete/move it. Therefore it might be a good idea to do this task when system activity is low or non-existent.
2 things also to check:
1. Make sure the python trace is not on.
2. In the python TREXAdmin Tool, check the Alerts tab and click "Alert Server Configuration". Make sure the trace level is set to "error".
Hope that helps. As always check the TOM for any concerns:
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/46e11c10-0e01-0010-1182-b02db2e8bafb
Edited by: Mike Bestvina on Apr 1, 2008 3:59 AM - revised some statements to be more clear

Report Server Generating big size trace file

We are using Report Service which is part of 9iAS suite. Its generating a trace file (server name.trc). This file keeps growing and depending on the number of requests we receive sometimes its size goes in Giga bytes. Anybody knows why its doing that and how to get rid of this behavior ?

Tracing is useful for troubleshooting purposes. If you do not need tracing please open the file
$OH\reports\conf\server_name.conf
and comment the "trace" element. You will need to restart the server for this to take effect

Thanks
The Oracle Reports Team

1250737 - SMMS: Trace file dev_ms could not be opened

Hi All,
"1250737 - SMMS: Trace file dev_ms could not be opened"
I have this error in smms on a CRM 7.0 W2k8 Failover Cluster, but the snote does not apply due to the version. Everything works, so I think it's not Security related.
Could you please help me in finding the root cause for the above error.
Thanks,
Rudolf

Hi Chetan,
according to smms (goto/Trace level/increase, decrease) it seems, that he trace level could be changed in every combination where CI/DI/ASCS are running (see investigations below):
But by chance I found the following (1.):
1. If only DI is started I can read the dev_ms via smms, when ASCS was on the "DI"-Node -> neither (I)* nor (II)*
- Moving ASCS to the "CI"-Node, ms_dev was not found -> (II)*
*Compared with Note 1250737:
(I) using CI -> dev_ms could not be opened (due to M in CI "dev_ms on C":)
(II) using Application-Server not on the "ASCS-Node" -> dev_ms not found (absolute Path used instead of share)
2. If only CI was started, dev_ms could not be opened, regardless where the ASCS is running -> (I)
3. Both Application-Server running and ASCS on "CI"-Node:
- logged on DI for smms trying reading dev_ms via DI -> dev_ms not found (II)
(logged on DI for smms trying reading dev_ms via CI -> dev_ms not found (II))
- logged on CI for smms trying reading dev_ms via DI -> dev_ms could not be opened (I)
(logged on CI for smms trying reading dev_ms via CI -> dev_ms could not be opened (I))
4. Both Application-Server running and ASCS on "DI"-Node:
- logged on DI for smms trying reading dev_ms via DI -> dev_ms could not be opened (?)
(logged on DI for smms trying reading dev_ms via CI -> dev_ms could not be opened (?))
- logged on CI for smms trying reading dev_ms via DI -> dev_ms could not be opened (I)
(logged on CI for smms trying reading dev_ms via CI -> dev_ms could not be opened (I))
I don't know if it's really related to note 125073, maybe it has something to do with access violation: why is (1.) working, and (4.) not?
Thank you very much for your effort
Regards
Rudolf

Agent10g: Size of Management Agent Log and Trace Files get oversize ...

Hi,
I have the following problem:
I had installed the EM Agent10g (v10.2.0.4) on each of all my Oracle servers. I've done a long time ago (a few months or a few years depending on wich server it was installed). Recently, I've got PERL error because the "trace" file of the Agent was too big (the emagent.trc was more than 1 Gb) !!!
I don't know why. I checked on a particular server on the AGENT_HOME\sysman\config (Windows) for the emd.properties file.
The following properties are specified in the emd.properties file:
LogFileMaxSize=4096
LogFileMaxRolls=4
TrcFileMaxSize=4096
TrcFileMaxRolls=4
This file had never been modified (those properties correspond to the default value). It's the same situation for all Agent10g (setup) on all of the Oracle Server.
Any idea ?
NOTE: The Agent is stopped and started weekly ...
Thank's
Yves

Why don't you truncate the trace file weekly. You can also delete the file. The file will be created automatically whenever there is a trace.

File Size capped at 32bit for File System Data soruce?

I have a report that uses the "File System Data" source. I am using it to find files older then 30 days of type BAK or TRAN. That all works well but the "File Size" field does not display the correct information. For example I have a file that is 191GB or 205,192,528,384 Bytes, but the report displays it as 4,294,967,295 this corresponds with the max for a 32bit INT. Can anyone confirm that this is a limitation of the driver? Is there are 64bit CRDB_FILESYSTEM.DLL?
Regards

Hi Thomas
What version of CR are you using? Please look for the version in the Help | About screen of the designer.
- Ludek
Senior Support Engineer AGS Product Support, Global Support Center Canada
Follow me on
Twitter

Delete TREX Trace Files and Turn TREX Off

Hi ,
We have massive amounts of TREX trace files that are taking up all of the memory. How do I delete these files? How do I know what files to delete? Also we are not using TREX in our BI server, how do I turn this OFF so that these files and never created?
Note from Moderator- don't use all caps in title as it is seen as shouting
Thanks
Edited by: Marilyn Pratt on Jan 27, 2009 12:19 PM

Hi Sanam,
For the TREX Preprocessor trace, You define the maximum file size for the input and output of documents in INI file - TREXPreprocessor.ini
The default size is set to 'Unlimited'
Please check the following values maintained in the configuration file.
Section: [ httpclient ]
Parameter: max_content_length
Default: Unlimited
Also read,
http://help.sap.com/saphelp_sm32/helpdata/en/1d/8d49cc304dd843a3a4a23d74888939/content.htm
Hope this is helpful.
Regards,
Varun

How do I keep an Edge Animate banner ad file size small?

Most publishers cap the file size of banner ads at 30 KB to 50 KB. Is that possible using Edge? I know someone asked a question like this two years ago, and the reply was to use a CDN to serve the larger .js files. Is there a more direct approach now that Edge is no longer in preview? Thanks!

Hi SGPGirl,
We were just forming what became the Edge Animate team two years ago, so many questions were asked at that time. The 30k-50k ad standards were built with Flash in mind; HTML advertising standards are still emerging. In the meanwhile, we are making a continuing investment to improve the form factor of Animate output. Here are a few things we've done so far in "Publish":
- Frameworks via CDN (for jQuery)
- "Polite" preloader - this allows you to sneak-in under 30k size limits. It shows a light-weight "preloader" and then pull in a larger payload after the page has loaded. The preloader UI contains a meter that indicates the gzipped size of the preloader including its resources.
- Compressed composition output. When published, the animation (xyz_edge.js) is written in an efficient/compact form that shaves as much as 80% from the composition size.
I'd be grateful for your perspective on the "polite preloader" approach vs. trying to fit the entire composition into 30k. If Animate were to support an "areo-gel weight" output format, what kinds of experiences would you expect to produce with it? If you have any examples, please share!
- Josh

Capping dev_ms trace file size

Similar Messages

Maybe you are looking for