You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Quite often we setup our config for LOGDNA_LOG_DIRS in linux (ubuntu) to monitor only a subset of folders, some of which don't exist at the time logdna-agent is started but will in the future. E.g a dated scenario:
And if a directory doesn't exist logdna will do a warning and I/O error (see below) and the directory will not get monitored even if it will eventually exist and be populated with logfiles.
[2022-08-30T23:45:01Z WARN config] /logs/weatherrun/2022/08/31/ is not a valid directory I/O error: No such file or directory (os error 2)
I understand the arguement of monitoring the parent folder instead, but due the volume of logfiles we generate this feels like a bad idea trying to monitor 1000s at once, and especially those which will no longer be updated since the day is over.
Is there/could a option be set for the agent to pick up directories its told to monitor even if they dont exist yet, I believe the v1 agent had this behaviour before?
The text was updated successfully, but these errors were encountered:
Piponater
changed the title
Agent monitoring, directories that dont existing at start time.
Agent monitoring directories that dont existing at start time.
Aug 31, 2022
Quite often we setup our config for LOGDNA_LOG_DIRS in linux (ubuntu) to monitor only a subset of folders, some of which don't exist at the time logdna-agent is started but will in the future. E.g a dated scenario:
LOGDNA_LOG_DIRS=/logs/2022/08/30,/logs/weatherrun/2022/08/31,/logs/2022/09/01
And if a directory doesn't exist logdna will do a warning and I/O error (see below) and the directory will not get monitored even if it will eventually exist and be populated with logfiles.
[2022-08-30T23:45:01Z WARN config] /logs/weatherrun/2022/08/31/ is not a valid directory I/O error: No such file or directory (os error 2)
I understand the arguement of monitoring the parent folder instead, but due the volume of logfiles we generate this feels like a bad idea trying to monitor 1000s at once, and especially those which will no longer be updated since the day is over.
Is there/could a option be set for the agent to pick up directories its told to monitor even if they dont exist yet, I believe the v1 agent had this behaviour before?
The text was updated successfully, but these errors were encountered: