hello friends, I come here to warn you that the version 5rc1 torch has a bug in the part that shows the TX is always at 0, since the RX ta working ok …
works fine for me:
show us yours, make a screenshot and post here
Normis, yes it works on base installation (installed on VMware) but not on production router.
Kamil
Image from picture is on “production router” so it works for me 
post screenshots of your problem
Uploaded with ImageShack.us
my router is x86… bug continua… v5rc1 bug torch x86, v5b6 not bug torch work fine…
TOTALLY UPD: yes, I can confirm such behaviour on x86. RBs don’t repeat that =)
It shows your CPU usage is %100
maybe it is the causes
plz retry after reboot
could be related to multicore on x86. we are investigating.
That is not a problem. The time when I made screenshot, through router was going heavy ipsec (15Mbps) traffic
Kamil
through router was going heavy ipsec (15Mbps) traffic
I’m suprised, your screenshot shows RB750.
What is your IPSec and Ethernet ports config?
Regards, Grzegorz.
md5, des.
What kind of throughput are you getting with 3DES and AES? Just curious how the different algorithms fare on that platform, and whether 3DES is more or less one third throughput and how AES compares to that.
md5,null - 20Mbps
md5,des - 12.5Mbps
md5,3des - 6Mbps
md5,aes-128 - 15Mbps
md5,aes-192 - 14Mbps
md5,aes-256 - 13Mbps
sha1,null - 2.5Mbps
sha1,des - 2Mbps
sha1,3des - 1.5Mbps
sha1,aes-128 - 2Mbps
sha1,aes-192 - 2Mbps
sha1,aes-256 - 2Mbps
null,null - not working
null,des - 13Mbps
null,3ds - 5Mbps
null,aes-128 - 18Mbps
null,aes-192 - 16Mbps
null,aes-256 - 14.5Mbps
all traffic were comming to RB750 from x86 machine
Wow, thanks for that. I didn’t expect sha1 vs md5 to make such a huge difference. sha1 is generally slower, but not by a factor of 5. Running checksums on 50MB test file with random data with OpenSSL (so disk IO routines etc are probably the same) have sha1 about 50% slower. Not that that’s a particularly scientific test, but still.
Thanks a lot !!!
Karma up !!!
Regards, Grzegorz.
I have another problem 
I confirm I saw this on 5.0b6 as well the tx rx on x86 wit 82576 controller. Same problem with rc1 it shows actual traffic in torch for multiple ip addresses so this is some kind of the bug. Its good to see other people also experiencing this. Using dell r200
I can confirm the multiple IP traffic torch bug. And can reproduce it easy. Was hoping it would be fixed in rc1 