Cache usage Check script under Linux system

The performance impact of the Linux cache Pagecache is critical, but how well we use it in the actual system, especially with regard to the utilization of each device.

From the following figure we can see very clearly:

We know that IO request is initiated by the VFS, after Pagecache cache, can not block the implementation of the IO device, then statistics this utilization is very simple. All we need to know is the proportion of Io that can't be blocked.

I wrote a systemtap script to solve this problem:

The code is as follows

Copy Code

$ uname-r 2.6.18-164.el5 $ sudo rpm-i kernel-debuginfo-common-2.6.18-164.el5.x86_64.rpm $ sudo rpm-i kernel-debuginfo-2.6.18-164.el5.x86_64.rpm $ cat >PCMR.STP Global Io_stat, Blk_stat Global rd_cnt=2, wr_cnt=3 Probe Vfs.read.return { if ($return >0 && devname!= "N/a") { Io_stat[devname, 0] + + $return Io_stat[devname, rd_cnt] + +; } } Probe Ioblock.request { Blk_stat[devname, rw] = = Size Blk_stat[devname, rw+rd_cnt] + +; } Probe Vfs.write.return { if ($return >0 && devname!= "N/a") { Io_stat[devname, 1] + + $return Io_stat[devname, wr_cnt] + +; } } Global _IO,_BLK Probe timer.ms (5000) { foreach ([Devname, Action] in Io_stat) { _io[action] + = io_stat[devname, action]; _blk[action] + = blk_stat[devname, action] } if (_io[0] + _io[1]) { printf ("n%-25s,%s%db/%d,%s%db/%d,%s%db/%d,%s%db/%d,%s%d%%nn", CTime (gettimeofday_s ()), "IO Read:", _io[0], _io[rd_cnt], "IO Write:", _io[1],_io[wr_cnt], "Blk Read:", _blk[0], _blk[rd_cnt], "Blk Write:", _blk[1],_blk[wr_cnt], "PCMR:", (_blk[0]+_blk[1]) */(_io[0] + _io[1))         Delete _io         Delete _blk        /* Print header */     printf ("%8s%2s%19s%19s%5sn",            "DEVICE", "T", "IO", "BLK", "PCMR")  } /* Print Top ten I/O foreach ([Devname, Action] in Io_stat-limit 10) if (Action < rd_cnt) printf ("%8s%2s%11db/%6d%11db/%6d%4d%%n", Devname, action? " W ":" R ", Io_stat[devname, Action],io_stat[devname, action+rd_cnt], Blk_stat[devname, Action],blk_stat[devname, action+rd_cnt], (Blk_stat[devname, Action] *)/io_stat[devname, action] ) /* Clear Data */ Delete Io_stat Delete Blk_stat } Probe begin { println ("::"); }

Ctrl+d

Arguments let's explain the following:
IO read:vfs level See IO Read/count
Io write:vfs level see iOS Write/number
Blk read: block level see IO Read/Count
Blk write: Io written at block level see Number of times
Pcmr:pagecache Miss Rate =100 * Blk/io, because file reads and writes are in page, it may >100%
The first line is a summary statistic, printed every 5 seconds.
Summary: PCMR is as small as possible, indicating high cache utilization.