| .TH dcstat 8 "2016-02-09" "USER COMMANDS" |
| .SH NAME |
| dcstat \- Directory entry cache (dcache) stats. Uses Linux eBPF/bcc. |
| .SH SYNOPSIS |
| .B dcstat |
| [interval [count]] |
| .SH DESCRIPTION |
| The Linux directory entry cache (dcache) improves the performance of file and |
| directory name lookups. This tool provides per-second summary statistics of |
| dcache performance. |
| |
| This uses kernel dynamic tracing of kernel functions, lookup_fast() and |
| d_lookup(), which will need to be modified to match kernel changes. |
| |
| Since this uses BPF, only the root user can use this tool. |
| .SH REQUIREMENTS |
| CONFIG_BPF and bcc. |
| .SH EXAMPLES |
| .TP |
| Print summaries each second: |
| # |
| .B dcstat |
| .TP |
| Print output every five seconds, three times: |
| # |
| .B dcstat 5 3 |
| .SH FIELDS |
| .TP |
| REFS/s |
| Number dcache lookups (references) per second. |
| .TP |
| SLOW/s |
| Number of dcache lookups that failed the lookup_fast() path and executed the |
| lookup_slow() path instead. |
| .TP |
| MISS/s |
| Number of dcache misses (failed both fast and slow lookups). |
| .TP |
| HIT% |
| Percentage of dcache hits over total references. |
| .SH OVERHEAD |
| The overhead depends on the frequency of file and directory name lookups. |
| While the per-event overhead is low, some applications may make over 100k |
| lookups per second, and the low per-event overhead will begin to add up, and |
| could begin to be measurable (over 10% CPU usage). Measure in a test |
| environment. |
| .SH SOURCE |
| This is from bcc. |
| .IP |
| https://github.com/iovisor/bcc |
| .PP |
| Also look in the bcc distribution for a companion _examples.txt file containing |
| example usage, output, and commentary for this tool. |
| .SH OS |
| Linux |
| .SH STABILITY |
| Unstable - in development. |
| .SH AUTHOR |
| Brendan Gregg |
| .SH SEE ALSO |
| dcsnoop(8) |
