m0unds_ haha
I was wanting to know if there is a way to find out why my computer is acting like it is on dailup when I am on cable and getting my full speed
that's a very open ended question
is dns resolution slow?
is it windows or linux or os x or what
basically i'd check ping first, then check dns, then check cpu usage, then check hard-disk performance
cdablre
It is cable
windows
ping is 0
Also what does this have to do with ARP Networks?
brycec: nothing really, but nothign else ios happening
but yeah it's probably a bit too far off-topic, and not generally intersting or anything
Hello
yeah mercutio, my thoughts exactly.
i've been playing with zfs on ssd
is that general interest?
it's still slower than btrfs or ext4 for read speeds :(
Under what OS?
linux
i got 3x samsung evo 840
So could just be the Linux ZFS implementation :P
could be
(Fuse, right?)
nope
it went past fuse ages ago
ah
i'm at about 1.2 gigabytes/sec though
and i got it up to 1.4 gigabytse/sec before
but ext4/btrfs on mdadm or on btrfs raid got up to 1.5 to 1.6gb/sec read speed
That seems faster than the bus should handle.
bus should handle 2 gigabytes/sec
minus overheads
@wa 6gbps to gb/s
convert 6 Gb\/s  (gigabits per second) to gigabytes per second;0.75 GB\/s  (gigabytes per second);6×10^9 b\/s  (bits per second);750 MB\/s  (megabytes per second);= SATA revision 3.0 speed ( 6 Gb\/s );~~ 1.2 × OC96 speed ( 4.977×10^9 b\/s );~~ 1.2 × USB 3.0 speed ( 4800 Mb\/s );information rate;[time]^(-1) [information];Time to transfer 1 gigabyte:,  ->1.3 seconds,  ->0.022 minutes;Time to transfer 1 terabyte:,  ->1333 seconds,  ->22 minutes
^
this is 3 ssd's
each on 6 gigabit ports
Ahhh so 3*.75GB/s
each ssd can do 500mb/sec+ on it's own
*its :p
it's hard to benchmark though
being evos they have fast write at the beginning but not after writing more data
So secure erase every time?
but in real world access you're not likely to write more than the fast speed often aynway
nah i'm not doing anything liek taht :/
i was mostly concerned about read speed aynway
heh
but then i decided to compare linux kernel compile time
I've done that before :)
and other such things
curiously lz4 is faster than non lz4
even on amazingly fast ssd's
doing 100% cpu
but basically zfs is slightly slower
19s was my record (lvm mirror with 2 SATA3 SSDs that max at 500-600 read, -j12)
but < 1 % with lz4 compression
and just over 1% without lz4 compression
19s to do what exactly?
make defconfig && make
wow
what cpu speed was that?
my kernel compile times went up at some point
Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz
and 16Gb
but it's like 2:07 for zfs
*16GB RAM
what
i'm on i7-4770
what does defconfig enable?
i'm using my standard kernel
i used to be just under 2 minutes
but yeh if it's minor difference it doesn't really matter
and lz4 uses up less space :)
No idea really... Just "defaults" as determined by Linus and firends.
and it seems raidz gives good enough performance
*friends
so i don't really need to use raid 10
because most things on linux don't do parallel i/o
and raid10 doesn't really speed up serial i/o
it just means that it doesn't slow down as much when doing parallel i/o
but ssd's don't really slow down that much with parallel i/o anytway
maybe it was an old kernel
i used to have 15 minute compile times on a 486 i think
Wow really? My 486DX2 took 2 hours
maybe i was imaginging things
(compiling some old 2.2 kernel, IDE drive...)
weird
oh maybe it was 40 minutes, and then 15 minutes when i
oh hangon
this was 2.1 kernel
like 2.1.110
so that looks like crica 98
My 486 died when I moved :(
damn
i going to see if it still compiles :/
it's not much gbigigger than current patches :/
Good luck :)
Makefile:160: arch/x86_64/Makefile: No such file or directory
ho hmm
% ls -l arch/
alpha/    arm/      i386/     m68k/     mips/     ppc/      sparc/    sparc64/
i need a 32 bit distro
i'm so used to using x64 arch that when i had to reimage something with a 32-bit cpu, i didn't have a compatible image for it
i found some old distribution
i'm hoping i can extract binaries from it :)
haha, good luck
bin/zsh: Linux/i386 demand-paged executable (QMAGIC)
doesn't seem to want to chroot and run that :/
i assume that's a.out