Open sebthebert opened 10 years ago
Date: 2012-11-14 10:34:31 Author: hollstein
screen shots, crontab and scripts
Date: 2012-11-14 10:35:15 Author: hollstein
Date: 2012-11-15 09:32:30 Author: hollstein
Hello,
since SF doesn't allow me to upload more than one attachment I uploaded another two images to my server. As you can see from these I disabled the cron jobs (octo-sanity-support) and the logs started dying down. Therefore there is of coure a corellation between DX and LX, since reloads cause CPU-usage. However the problem is not the CPU usage which I can handle, but the starvation of logs within the RRD-Graph. Maybe I should have a look at the raw logs and see whether less log data is stored on disks over time? :-]
URL: tron-delta.org/octo-nagios-croellation-dw-2012-11-14-sls-cn.tar.bz2
Date: 2012-11-16 10:44:09 Author: hollstein
Should I file a bug report on this? do you need further data to assist you in analyzing the problem?
I added this as a feature request, but not as a bg since I don\'t know whether this is one. Further I selected category \"Dispatcher\" but not \"Interface Improvements\", since I think the dispatcher is pretty much involved in this.
For this feature request (a help request actually) I added some files contained within an archive. These are basically screen shots, crontab and scripts for you to verify.
The problem I have/the feature I request is basically that:
Logs are not dying down over time so a restart of Octopussy via crontab is not required anymore. As seen from the monthly graph that happens and I don\'t know why.
Octopussy to (somehow) buffer logs until completely restarted. Maybe we should implement the concept of different softwares like Apache2 with \"reload\" and \"restart\" where reload does not kill a main process.
The RRD graph tool (I just had a quick look at the code) may have to be altered to satisfy the aforementioned needs.
I hope there is an easy solution to this. Maybe I\'m interpreting the data wrong, but maybe logs get lost within the system which demands those restarts every 8 hours or the RRD graph tool simply misbehaves. In case you need more data please tell me and I can send you nearly every file (code and configuration) from our Octopussy installation.