Thanks, diff applied to both master and 2-2-1 branches
ticket #334 Put back connection errors on stderr
ticket #334 Put back connection errors on stderr
Put back connection errors on stderr
remove CCA and reverse restriction
connection report without enhanced needs to be null for timestamp
Move all read/writes including cmsg ones into socket_io code
Out of order seq no
fix missing parenthesis
fix ip_tos cmsg
use ancillary and sendmsg on client side l4s
move memset
remove setsockopt for ip tos, use sendmsg and ancillary messages instead
use ancillary cmsg and sendmsg to set L4S tos (ecn)
remove superfulous mysubUsec method
remove superfulous mysubUsec method
update prague files per l4s git hub
fix out of order with --udp-l4s
separate mBuf from ack buf on l4s server side
minor fix to ack_poll
add l4s in connection report
fix segfault when device is given on the client
remove why ending debug
l4s ack poll routine
fixe to dist-all
use poll in place of select
Merge branch 'master' into udp_prague
fix cca buffer overflow in connection report code
Ok, I think I found the buffer overflow. It's in the CCA parts. Hopefully fixed now.
fix to other cca buffer overflows
hopefully fix buffer overflow in connect message (in CCA append), add negative test for snprintf
Thanks for this. Can you try master now? 1. Code commit to protect against connect report buffer overflows (in master) 2. Code commit for ifndef I can't speak to low performance. My systems drive 10G links fine. Let me know what you figure out or if I can help diagnose somehow.
Thanks for this. Can you try master now? 1. Code commit to protect against buffer overflows in master 2. Code commit for ifndef I can't speak to low performance. My systems drive 10G links fine. Let me know what you figure out or if I can help diagnose somehow.
Thanks for this. Can you try master now? 1. Code commit to protect against buffer overflows in master 2. Code commit for ifndef https://sourceforge.net/p/iperf2/code/ci/756dbf80c40298f67a513425e6ee739af6f1cf07/ I can't speak to low performance. My systems drive 10G links fine. Let me know what you figure out or if I can help diagnose somehow.
bump date
add ifndef for IPTOS_DSCP_NQB and DQB2 before define not to overload
add ifndef for IPTOS_DSCP_NQB and DQB2 before define not to overload
fail on buffer overflow in connection report
don't use a separate thread for udp acks
Merge branch 'udp_prague' of ssh://git.code.sf.net/p/iperf2/code into udp_prague
more L4S coding
init code for separate udp ack thread
set reserved to zero
set reserved to zero
Merge branch 'udp_prague' of ssh://git.code.sf.net/p/iperf2/code into udp_prague
l4s proto moving packets
add thread debug into L4S startup
udp prague compiles, no testing
sender side l4s send
add some field commnets to L4S ack packet
add brahch name udp_prague to version output
working l4s sender side
server side l4s code compiles, no testing
add l4s payload fields
Merge branch 'master' into udp_prague
udp shouldn't use myWrite decorator but call write directly
udp shouldn't use myWrite decorator but call write directly
udp shouldn't use myWrite decorator but call write directly
udp shouldn't use myWrite decorator but call write directly
add prauge pacer object
add L4S run loop on server side
more indent fixes
more indent fixes
more indent fixes
more indent fixes
more indent fixes
fix indenting
pass L4S test flag, support IP_RECVTOS for L4S thread
a bit more on L4S
remove blank line
udp prague initial code
Can you file a ticket for this? Iperf should pr int the error code but it currently doesn't.
udp prague inital files
Hi Andreas, Are you asking the server side output to be on the client side too? Currently the server side output is only on the server side and the client side output is on client side.
make sure to get at least one tcp sample per an interval
make sure to get at least one tcp sample per an interval
bump date
bump date
Merge branch 'master' into 2-2-1
Merge branch 'master' of ssh://git.code.sf.net/p/iperf2/code
sample tcp_info in the middle of a -i interval
fix windows compile
Revert "add byte wait times to outputs when inflight bytes are available"
Revert "improve inflight & wait_time ifdef"
sample tcp_info ahead of the write so the stack has time to run before taking the sample
tcp_info sampling at the wrong time
On bytes in flight per a BBR discussion
On bytes in flight per a BBR discussion
Hi All, FYI, here is a good review of Little's Law applied to packets. A major challenge our industry faces is making responsiveness KPIs more prevalent and actionable to network engineers. Currently, many think of capacity only metrics to qualify a link. I'm considering adding the time to service a byte to iperf 2 since TCP is a byte oriented protocol. This time will be calculated on the sender side using Little's law. The "depth" is taken from the tcp_info bytes in flight calculation and the arrival...
Hi All, FYI, here is a good review of Little's Law applied to packets. A major challenge our industry faces is making responsiveness KPIs more prevalent and actionable to network engineers. Currently, many think of capacity only metrics to qualify a link. I'm considering adding the time to service a byte to iperf 2 since TCP is a byte oriented protocol. This time will be calculated on the sender side using Little's law. The "depth" is taken from the tcp_info bytes in flight calculation and the arrival...
Hi All, Note that there are engineers from multiple companies on this email. Feel free to engage others or add them to the cc list, but keep the information shared limited to the public domain. FYI, here is a good review of Little's Law applied to packets. A major challenge our industry faces is making responsiveness KPIs more prevalent and actionable to network engineers. Currently, many think of capacity only metrics to qualify a link. I'm considering adding the time to service a byte to iperf...
add files to doc directory in tarball (or make dist-all)
add files to doc directory in tarball (or make dist-all)
improve inflight & wait_time ifdef
add byte wait times to outputs when inflight bytes are available
don't print out settings report for working-load
don't print out settings report for working-load
check for TCP_TX_DELAY as a socket option
check for TCP_TX_DELAY as a socket option
fixes for HAVE_DECL_TCP_TX_DELAY and compile breakages