monitorio.sh eats too much I/O!

[mudkip]

I have millions of small files (backup of my website) stored in one share of My Book Live. I notice green led flicker too often after I stored these files. And the speed is much lower than before.

So I sshed onto My Book Live and I found something terrible. Tthe %wa(which means waiting for I/O) keeps being very high. And the criminal is "ls, tally and monitorio.sh". So I looked monitorio.sh and there's a "ls -s1NRA --block-size=1 /shares", this command list all files on my shares and it seems taking hours to traverse the whole drive.

Is there any way to change this?

[mudkip]

I moved these small files out of /shares and to /DataVolume/backup and it's OK now, and I can delete the annoying ".AppleDouble" folders. Though I can only access these files via SFTP.

[stefan29]

Hello,

same problem here. Only this happens after each (incremental) Time Machine backup, since TM backups are sparsebundles consisting of thousands of 8MB files. Also, I have tens of thousands of files in my /shares directory.

Can this job be disabled safely? (I can do it via SSH if it doesn't break anything.)

Thanks!