Showing results for 
Show  only  | Search instead for 
Did you mean: 

Limit for automatic trace recording


We are trying to set up automatic traffic recording via the RUM Console.

We can limit the time for each trace and the delay between two traces.

Is there any way to limit the disk space used for this traces? We want to take care that the disk doesn't fill up with trace recordings.



Hi Friederike

I might be out on thin ice here but I think there is some sort of clean up process.

I have a rather small AMD in my lab and it has the recorder running the whole time. Yet there are never a lot of files so my guess is that there is something removing them. Perhaps the lab has some ideas? @Sebastian Kruk @Krzysztof Ziemianowicz

Dynatrace Pro
Dynatrace Pro

If I recall correctly, this will be automatically cleaned to maintain a minimum of 20% free space on the filesystem.

-- Erik

Yes - that would explain a few things 🙂

But it would be even better if that was a fact, like written somewhere.

As things are now, I often find customers that has concerns about disk space on the AMD.

Dynatrace Pro
Dynatrace Pro

Erik is right. The file is called datacleaner.config and it is located in the usr/adlex/config directory.

The parameter minFreeSize.percent = 20 specifies that the AMD maintains a minimum of 20% disk space.



Dynatrace Pro
Dynatrace Pro

/usr/adlex/config/datacleaner.config covers /var/spool/adlex/rtm/ while the automatic traces are in /var/spool/adlex/cba/

doing further research, I found this segment in /usr/adlex/config/cba.config.xml


updated information:

"autoSampleThreshold" is in hours, and any auto samples that are older than the specified limit are purged. This is applied to /var/spool/adlex/cba/auto/ and cycles every new auto-capture attempt.

"minDiskSpace" is a the required free percentage of the partition's total, and the oldest *.cap file from /var/spool/adlex/cba/auto/ anytime free space falls below this percentage. This is also applied/tested every new auato-capture cycle. If the threshold is breached and purging old *.cap files can't resolve it, all new capture attempts automatically fail.

"samplesMaxDiskAllocation" is in GB, and applied to the /var/spool/adlex/cba/ directory; if the total size of all *.cap files in the directory tree is greater than the allocation limit, all new capture attempts automatically fail.

(credit for the updated information goes to Dev)

-- Erik


first of all thanks to all of you for the research.

I found the same file like Erik but I was wondering about the units and if this really adresses automatic traces. I don't wanted to do some test with this file in the productive environment.