During a dump(8) of a filesystem on da0 to a holding file on da1,
with nothing else going on, iostat(8) reports figures such as these:
da0 da1 cd0 cd1
KB/t tps MB/s KB/t tps MB/s KB/t tps MB/s KB/t tps MB/s
64.00 148 9.28 64.00 24 1.52 0.00 0 0.00 0.00 0 0.00
64.00 134 8.38 64.00 17 1.07 0.00 0 0.00 0.00 0 0.00
63.63 134 8.32 64.00 18 1.10 0.00 0 0.00 0.00 0 0.00
64.00 143 8.96 60.30 24 1.41 0.00 0 0.00 0.00 0 0.00
64.00 140 8.73 63.85 21 1.33 0.00 0 0.00 0.00 0 0.00
64.00 130 8.14 62.13 15 0.93 0.00 0 0.00 0.00 0 0.00
64.00 138 8.63 64.00 18 1.10 0.00 0 0.00 0.00 0 0.00
63.79 137 8.56 64.00 18 1.10 0.00 0 0.00 0.00 0 0.00
63.82 140 8.74 64.00 18 1.10 0.00 0 0.00 0.00 0 0.00
64.00 138 8.61 61.67 24 1.47 0.00 0 0.00 0.00 0 0.00
Why is the I/O ratio so skewed towards reading? I would have
expected the I/O on the reading side to mostly match the writing
side. Instead, several times as much data appears to be read as
is written.
And yes, this uses dump's recently introduced -C cache option.
--
Christian "naddy" Weisgerber [EMAIL PROTECTED]
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-questions" in the body of the message
