Solaris

Hi,

I would like to is it possible to repair inode table in solaris without running fsck? I am facing this issue in root file system, because to run fsck I have to bring system in single user mode with booting cdrom media.

I facing strange issue in /var/adm/messages it is showing inode table full , but while running df command it is showing file system free space.

Please suggest me how to resolve this issue?

Run the below command to know the inodes status of root FS.
df -o i /

It appears that all the inodes are used, and no more files can be created.

Free space is available because of the data blocks which are still there to be linked to the existing inodes when a file needs to be grown.

Can you post the exact error message, the matching line from /etc/fstab and "df -ki" for the filesystem in question.

Hi Vishal,

Thanks for your quick help i have checked the output of above command it was showing inode full. after that I have deleted the files, but still issue persists.

Regards,

Manoj

Please post "df -i" output.
Which files did you delete ?
It might be a process that creates new files constantly.

Please post sample error messages from /var/adm/messages and any error messages you see. Don't paraphrase - just post the exact error message.

Remember that there is one inode table for every mounted filesystem. This is why people are asking to see the output from various "df -i" commands. If you need to save the output to a file it must of course be saved to a filesystem which has free inodes. If you are having trouble with say the "more" command this suggests that "/tmp" is affected.
If you are getting random failures across multiple filesystems this could be a kernel parameter issue.

I am getting below messages in /var/adm/messages.

ID 682040 kern.notice] NOTICE: /: out of inodes

Please suggest.