June 2003 - Udpcast

Problems with Multiple Switches...
by Saikrishna Dasari 25 Jun '03

25 Jun '03

Hi Alian, I was trying to clone a 12 GB HDD on a dell machine to another machine of similar architecture but both of these are connected via multiple switches( 3 ) and these machines belong to the same network 192.168.17.0 I was able to do this cloning when there was only one switch between these 2 but when there are more than one switch UDPCast fails with Timeout problems and gives a message on the sender side that it is dropping client bacause of timeout. I am using Centercom FS708 swicthes in all the cases described above. Why is it failing for multiple switches?? is there are any workaround for this. Thanks in advance, Sai --------------------------------- Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month!

1 0

Default updcast clone session fails at 2.54 gigs
by Jim Christiansen 23 Jun '03

23 Jun '03

Hello, I may be having network problems but I'm not sure. When the transfer gets to 2.254 gigs the transfer fails showing many send errors then finally says, always at 2.254 gigs "send error". I am on a 100 megabit network using computers with 10/100 realtech 8139s. The updcast software brings up 2 nic modules and the one that works is called ...8139too. Some 10 megabit devices are also running on it. The cd is very easy to use using the default setting. I tested the udpcast software at home on a fully 100 mega bit network and everything just worked fine. I guess (or hope that) the problem must be with our old network at school that has many old devices hanging on it or old and bad wiring. How should I use the --half-duplex option and would this help? Also, I tried an older crossover cable and it only ran at 30 megabit or so and also failed going directly from machine to machine at the 2.254 gig mark. It showed the same errors the busy network transfer. It worked at home, so it must work at school... Thanks for any ideas, Jim _________________________________________________________________ STOP MORE SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=features/junkmail

1 0

Timeout notAnswered=[...]
by Patterson, Michael 20 Jun '03

20 Jun '03

Hello, I have installed the 20030607 version of udpcast. Udpcast looks like a great application, but I am having mixed success with it. I'm attempting to multicast files within a 100 Mbit ethernet network to the servers of a 500+ node Beowulf. These nodes are connected by three class B subnets. Three Cisco switches (two are 6513s and one is a 6509) route the traffic across this network. Udpcast is working for files of arbitrary size (I've tested up to 100 Mbytes) when multicasting to 80 or fewer nodes. For larger numbers of nodes (say, 100) the transfer appears to procede correctly but the receiving nodes then have difficulty disconnecting from the sending node. The receiving nodes all eventually disconnect, but they can require several minutes to do so. I've tried various combinations of options but none have resolved this problem. Can anyone provide guidance here? The command lines I am using are as follows: sender: udp-sender --portbase ${PORT} --min-clients ${NODES} --nokbd -b 1024 --max-bitrate 10m --ttl 1 --full-duplex --mcast-addr ${IP_CAST} --file ${SOURCE_FILE} --log /tmp/udp.log --autostart ${P} receiver: udp-receiver --portbase ${PORT} --nokbd --ttl 1 --file ${DEST_FILE} &> /dev/null A sample output from the sender is listed below. Thanks, -Mike ........................................................... MC=x 0xbffff950 0xbffff970 Udp-sender 2003-06-11 UDP sender for /tmp/x.mat at 172.18.124.7 on eth0 Broadcasting control to 172.18.127.255 New connection from 172.18.127.99 (#0) 00000019 New connection from 172.18.127.96 (#1) 00000019 New connection from 172.18.127.92 (#2) 00000019 New connection from 172.18.127.95 (#3) 00000019 New connection from 172.18.127.100 (#4) 00000019 New connection from 172.18.127.90 (#5) 00000019 New connection from 172.18.127.97 (#6) 00000019 New connection from 172.18.127.93 (#7) 00000019 New connection from 172.18.127.89 (#8) 00000019 New connection from 172.18.127.87 (#9) 00000019 New connection from 172.18.127.88 (#10) 00000019 New connection from 172.18.127.94 (#11) 00000019 New connection from 172.18.127.64 (#12) 00000019 New connection from 172.18.127.67 (#13) 00000019 New connection from 172.18.127.91 (#14) 00000019 New connection from 172.18.127.70 (#15) 00000019 New connection from 172.18.127.65 (#16) 00000019 New connection from 172.18.127.56 (#17) 00000019 New connection from 172.18.127.45 (#18) 00000019 New connection from 172.18.127.51 (#19) 00000019 New connection from 172.18.127.59 (#20) 00000019 New connection from 172.18.127.12 (#21) 00000019 New connection from 172.18.127.39 (#22) 00000019 New connection from 172.18.127.50 (#23) 00000019 New connection from 172.18.127.13 (#24) 00000019 New connection from 172.18.127.33 (#25) 00000019 New connection from 172.18.127.74 (#26) 00000019 New connection from 172.18.127.63 (#27) 00000019 New connection from 172.18.127.85 (#28) 00000019 New connection from 172.18.127.72 (#29) 00000019 New connection from 172.18.127.38 (#30) 00000019 New connection from 172.18.127.36 (#31) 00000019 New connection from 172.18.127.98 (#32) 00000019 New connection from 172.18.127.52 (#33) 00000019 New connection from 172.18.127.77 (#34) 00000019 New connection from 172.18.127.66 (#35) 00000019 New connection from 172.18.127.48 (#36) 00000019 New connection from 172.18.127.73 (#37) 00000019 New connection from 172.18.127.58 (#38) 00000019 New connection from 172.18.127.41 (#39) 00000019 New connection from 172.18.127.34 (#40) 00000019 New connection from 172.18.127.40 (#41) 00000019 New connection from 172.18.127.27 (#42) 00000019 New connection from 172.18.127.46 (#43) 00000019 New connection from 172.18.127.102 (#44) 00000019 New connection from 172.18.127.18 (#45) 00000019 New connection from 172.18.127.78 (#46) 00000019 New connection from 172.18.127.37 (#47) 00000019 New connection from 172.18.127.28 (#48) 00000019 New connection from 172.18.127.55 (#49) 00000019 New connection from 172.18.127.22 (#50) 00000019 New connection from 172.18.127.44 (#51) 00000019 New connection from 172.18.127.17 (#52) 00000019 New connection from 172.18.127.57 (#53) 00000019 New connection from 172.18.127.20 (#54) 00000019 New connection from 172.18.127.47 (#55) 00000019 New connection from 172.18.127.68 (#56) 00000019 New connection from 172.18.127.54 (#57) 00000019 New connection from 172.18.127.62 (#58) 00000019 New connection from 172.18.127.42 (#59) 00000019 New connection from 172.18.127.26 (#60) 00000019 New connection from 172.18.127.80 (#61) 00000019 New connection from 172.18.127.71 (#62) 00000019 New connection from 172.18.127.83 (#63) 00000019 New connection from 172.18.127.25 (#64) 00000019 New connection from 172.18.127.81 (#65) 00000019 New connection from 172.18.127.35 (#66) 00000019 New connection from 172.18.127.29 (#67) 00000019 New connection from 172.18.127.23 (#68) 00000019 New connection from 172.18.127.43 (#69) 00000019 New connection from 172.18.127.49 (#70) 00000019 New connection from 172.18.127.21 (#71) 00000019 New connection from 172.18.127.19 (#72) 00000019 New connection from 172.18.127.15 (#73) 00000019 New connection from 172.18.127.53 (#74) 00000019 New connection from 172.18.127.86 (#75) 00000019 New connection from 172.18.127.30 (#76) 00000019 New connection from 172.18.127.60 (#77) 00000019 New connection from 172.18.127.32 (#78) 00000019 New connection from 172.18.127.24 (#79) 00000019 New connection from 172.18.127.82 (#80) 00000019 New connection from 172.18.127.31 (#81) 00000019 New connection from 172.18.127.79 (#82) 00000019 New connection from 172.18.127.75 (#83) 00000019 New connection from 172.18.127.14 (#84) 00000019 New connection from 172.18.127.84 (#85) 00000019 New connection from 172.18.127.16 (#86) 00000019 Starting transfer: 00000019 bytes= 8 000 184 re-xmits=000000 ( 0.0%) slice=0112 8 000 184 - 60 00000 ( 0.0%) slice=0112 8 000 184 - 80 Disconnecting #8 (172.18.127.89) Disconnecting #64 (172.18.127.25) Disconnecting #21 (172.18.127.12) Disconnecting #18 (172.18.127.45) Disconnecting #4 (172.18.127.100) Disconnecting #22 (172.18.127.39) Disconnecting #85 (172.18.127.84) Disconnecting #33 (172.18.127.52) Disconnecting #36 (172.18.127.48) Disconnecting #66 (172.18.127.35) Disconnecting #47 (172.18.127.37) Disconnecting #58 (172.18.127.62) Disconnecting #39 (172.18.127.41) Disconnecting #73 (172.18.127.15) Disconnecting #59 (172.18.127.42) Disconnecting #31 (172.18.127.36) Disconnecting #76 (172.18.127.30) Disconnecting #70 (172.18.127.49) Disconnecting #48 (172.18.127.28) Disconnecting #51 (172.18.127.44) Disconnecting #1 (172.18.127.96) Disconnecting #50 (172.18.127.22) Disconnecting #0 (172.18.127.99) Disconnecting #17 (172.18.127.56) Disconnecting #13 (172.18.127.67) Disconnecting #12 (172.18.127.64) Disconnecting #55 (172.18.127.47) Disconnecting #23 (172.18.127.50) Disconnecting #11 (172.18.127.94) Disconnecting #24 (172.18.127.13) Disconnecting #77 (172.18.127.60) Disconnecting #20 (172.18.127.59) Disconnecting #29 (172.18.127.72) Disconnecting #6 (172.18.127.97) Disconnecting #71 (172.18.127.21) Disconnecting #40 (172.18.127.34) Disconnecting #32 (172.18.127.98) Disconnecting #46 (172.18.127.78) Disconnecting #7 (172.18.127.93) Disconnecting #2 (172.18.127.92) Disconnecting #72 (172.18.127.19) Disconnecting #53 (172.18.127.57) Disconnecting #41 (172.18.127.40) Disconnecting #30 (172.18.127.38) Disconnecting #3 (172.18.127.95) Disconnecting #49 (172.18.127.55) Disconnecting #79 (172.18.127.24) Disconnecting #27 (172.18.127.63) Disconnecting #5 (172.18.127.90) Disconnecting #42 (172.18.127.27) Disconnecting #54 (172.18.127.20) Disconnecting #65 (172.18.127.81) Disconnecting #9 (172.18.127.87) Disconnecting #56 (172.18.127.68) Disconnecting #52 (172.18.127.17) Disconnecting #14 (172.18.127.91) Disconnecting #78 (172.18.127.32) Disconnecting #81 (172.18.127.31) Disconnecting #62 (172.18.127.71) Disconnecting #28 (172.18.127.85) Disconnecting #15 (172.18.127.70) Disconnecting #67 (172.18.127.29) Disconnecting #74 (172.18.127.53) Disconnecting #45 (172.18.127.18) Disconnecting #10 (172.18.127.88) Disconnecting #68 (172.18.127.23) Disconnecting #43 (172.18.127.46) Disconnecting #86 (172.18.127.16) Disconnecting #69 (172.18.127.43) Disconnecting #63 (172.18.127.83) Disconnecting #16 (172.18.127.65) Disconnecting #44 (172.18.127.102) Disconnecting #34 (172.18.127.77) Disconnecting #60 (172.18.127.26) Disconnecting #35 (172.18.127.66) Disconnecting #38 (172.18.127.58) Disconnecting #82 (172.18.127.79) Disconnecting #84 (172.18.127.14) Timeout notAnswered=[25,61] notReady=[25,61] nrAns=7 nrRead=7 nrPart=9 avg=12407 Timeout notAnswered=[25,61] notReady=[25,61] nrAns=7 nrRead=7 nrPart=9 avg=12407 Timeout notAnswered=[25,61] notReady=[25,61] nrAns=7 nrRead=7 nrPart=9 avg=12407 Timeout notAnswered=[25,61] notReady=[25,61] nrAns=7 nrRead=7 nrPart=9 avg=12407 This last line will repeat dozens to hundreds of times.

2 1

[Udpcast] UDPCast Performance : 1 Server, 1 Client without any switch
by Saikrishna Dasari 17 Jun '03

17 Jun '03

Hi Alian, I have been recording the performance statistics of UDPCast for different system setups ( i.e ..(1 server, 3 clients), ( 1 server, 8 clients), (1 server, 1 client) etc) . When I was testing the 1 Server , 1 Client configuration I connected the server and client with a direct cable inbetween and since those two machines are independent of other netwroks I was expecting no retransmission messages during the transfer..But UDPCast shows several retransmission messages . I was confused by this. Since there are only two machines that are directly hooked up by a cable there shouldn't be any network traffic or congestion related issues. and these two machines are exactly identical in every respect. I am curious to know the possible reason. Thanks in advance. Sai --------------------------------- Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month!

2 1

Timeout
by Andrew Cooks 17 Jun '03

17 Jun '03

Hi How long is the timeout on the new release? I used the new CD image (I think it's 20030607, or 20030610) and I found that I have to restart the operation if any of the receiver machines stop receiving. "Stop receiving" includes reboots and "No space left on device" errors. I know there was a timeout value that was sometimes too short and caused "Pipeline full" errors and consequent dropping of the clients, but I haven't had a chance to look at the source again. Andrew ----------------------- Andrew Cooks TechTeam TechTeam -- "We make it work." Computer Science dept. University of Pretoria ----------------------- Stimulate your mellon for a change! Use Linux! - http://tlug.up.ac.za "If 44,000 employees of Sun can work with StarOffice, and can exchange any document with their customers, there is no good argument not to do it" - Richard Seibt, CEO of Suse Linux

2 1

Image Distribution Consistency
by Gianni Minetti 15 Jun '03

15 Jun '03

Hi to everybody. I understand that udpcast guarantees the integrity of the transmitted data through FEC redundancy code model but how can one be assured , beside integrity, also of consistency of the transmitted data, defining consistency as the property the all the hosts being part of the multicast group have received the data? I mean, if my multicasting model is composed of N host and I schedule to send a different file every T minutes, how can I detect the condition of D hosts not reachable for a time >T (tahr result in data copied only on N-D hosts)? Should I check the 'health' of all my N hosts before to run udpcast using icmp? For the application I've in mind it's very important to know what has or hasn't been delivered to whom so, I would like to encourage everyone who has had a similar problem or has some idea how to cope with. Thanks a lot in advance for your suggestions. Regerds, Gianni

2 1

another e1000.o
by Nathaniel Sbar 15 Jun '03

15 Jun '03

Just out of curiousity, I have another e1000.o question. I hooked two pc's up to each other with a crossover cable (so no switch needed), and started UDPcasting from one to the other without compression. I was using the 20030524 build. Top speed on the transfer was about 92Mbps. That's great and all, but the NIC is a gigabit ethernet NIC and I was hoping that there was some way to multiply that speed by ten (give or take). Is this something a user can do, or is this something that takes talent (as in coding skill)? Nate Sbar. __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com

2 1

[Udpcast] Udpcast 20030601: support for integrating custom scripts and improved PCMCIA handling
by Alain Knaff 11 Jun '03

11 Jun '03

There is a new version, 20030601, which makes it easyer to include custom scripts to be run before or after the transfer. * Udpcdialog now has hooks that are called before and after the actual transfer: /udpreceiver.pre This script is invoked before starting udpreceiver /udpreceiver.post This script is invoked after udpreceiver has finished. It gets called with a single parameter, which is the exit status of udpreceiver /udpsender.pre This script is invoked before starting udpsender. /udpsender.post This script is invoked after udpsender has finished. It gets called with a single parameter, which is the exit status of udpsender * These scripts can easily be put into the initrd by using makeImage's new --merge option: makeImage -c cd.img -k /boot/vmlinuz-2.4.21-rc6-udpcast \ --merge /udpreceiver.pre=my-pre-script \ --merge /udpreceiver.post=my-post-script * There is now better support for PCMCIA cards (cardmgr included, automatic probing for "socket" driver modules). However, this is still largely untested. * lzop is now integrated in busybox, rather than being a statically compiled standalone binary. This saves some space. Have fun, Alain

2 9

Udpcast 20030609 stable released
by Alain Knaff 09 Jun '03

09 Jun '03

This is to announce that the new udpcast is now considered stable, and is directly available from the main page (http://udpcast.linux.lu) Recent changes include: - Fixed cardmgr's output, which messed up the dialog - Added reference to the BSD-like license of fec.c and cardmgr to the documentation, and added a new --license flag as per the "publicity clause" (only available on the standalone version. The embedded version doesn't use fec.c anyways). - XDF bootable floppy disk, for having an even larger choice of network modules on a single floppy disk (XDF can hold 1840 KB of data, and can thus fit 647K of kernel, and 1207 of initrd) Regards, Alain

1 0

is busybox a requirement?
by tomas pospisek 04 Jun '03

04 Jun '03

As is udpcast wants a busybox installation in order to compile. I've replaced all "libbb_udpcast.h" occurences by "udpcast.h" (my brain was in standby mode) and all "xmalloc" calls by "malloc" and udpcast did compile. I couldn't test it yet (no network at home), but it runs. Alain, do want to add a feature to compile without BB or should I send you a patch? *t -- Bad times for dictators. At least in the oil regions. Goedart Palm

1 0