DX Unified Infrastructure Management

 We are running int a situation , on UNIX and Linux systems, where a file system and/or a disk goes into Read-Only mode. 

 At this point, most system/application functionality will stop because the system can no longer write to a file, for example. Monitoring for this situation difficult. Once the systems gets into R/O mode, most often we cannot log in to the system in order to check anything, such as log outputs , to detect this R/O situation. 

 Does anyone have any ideas for a solution ?

Thanks.

David

If the mode change to read only is recorded in the system 'messages' or some other log, then logmon can monitor that log for the message string indicating the change. .

logmon (Log Monitoring) - CA Unified Infrastructure Management Probes - CA Technologies Documentation 

Hi dmdistel ,

Were you able to test logmon as per David's suggestion? Did it help?

Thanks in advance for your feedback

Octavie

Thank you to those who have offered suggestions.

I am currently working with my UNIX team on various options and testing functionality and results.

Thanks.

I would appreciate if you post your results and the solution you eventually come up with.  We need to do the same thing, but I have not started this task yet.

Thanks,

Chris

In mine scenario i use an indirect method, that is check the disk IO with the VMware probe, if the metric goes bellow the baseline it flags me an alarm

Use logmon to run a script.

In that script, use touch to create a file. Then test if the file exists (if -f) and has a current timestamp (find +mtime). Then remove the file.

If you find no file or a file with an old timestamp, the filesystem is read only.

-Garin

I found a article for you .

How can I monitor a read-only filesystem on a Linu - CA Knowledge