Return to NOTES CVS log

Up to [Research Unix] / researchv8dc / netfs

researchv8 Dan Cross

/tmp/xxx looks like a 1048576 byte file, but it is all holes except for the last
block.
:;time cat /tmp/xxx > /dev/null
0.1u 2.4s 3r 	 cat /tmp/xxx
:;time cat /n/debug/tmp/xxx > /dev/null
0.2u 20.0s 66r 	 cat /n/debug/tmp/xxx
Thus the netfs without disk delay is running at 15887 bytes per second.
If the file is buffered in the server there is no noticable change.
Thus the slow rate is in the protocol.
ws%time cat /n/seki/tmp/__2148 > /dev/null #totally buffered in server
0.1u 21.2s 52r 	 cat /n/seki/tmp/__2148
ws%time cat /n/seki/tmp/xxx > /dev/null #hole file, no buffering in server
0.0u 20.8s 53r 	 cat /n/seki/tmp/xxx
ws%time cat /n/seki/tmp/xxx > /dev/null #no holes either
0.0u 20.8s 53r 	 cat /n/seki/tmp/xxx
the netfs runs at bout 19784 bytes per second, with no perceptible
effect of disk or buffering at host end.

1000 local fstats take about 11 seconds of system time,
while 1000 getuids take about 2.7 seconds.
1000 remote fstats of /n/debug take 33 seconds of real time, about 10 milliseconds
a message.

Now if the server and setup are connected through a pipe, instead of the
network, catting the file with holes:
:;time cat m/tmp/9489 > /dev/null
0.1u 5.1s 12r 	 cat m/tmp/9489
without holes:
:;:;time cat m/tmp/x > /dev/null
0.1u 5.2s 12r 	 cat m/tmp/x

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.