planetlab-users Mailing List for PlanetLab (Page 7)
Brought to you by:
alklinga
Menu
▾
▴
planetlab-users — Discussion among PlanetLab users; but send trouble reports to planetlab-support
You can subscribe to this list here.
| 2002 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(3) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2003 |
Jan
(2) |
Feb
(2) |
Mar
(7) |
Apr
(6) |
May
(5) |
Jun
(8) |
Jul
|
Aug
(9) |
Sep
(18) |
Oct
(24) |
Nov
(46) |
Dec
(37) |
| 2004 |
Jan
(65) |
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
(2) |
Jul
(1) |
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
| 2005 |
Jan
(2) |
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Paul B. <pau...@pl...> - 2003-11-10 19:20:51
|
Massimo Are you associated with a current PlanetLab site, or are you planning to bring up nodes at UNIPMN? If you are planning to bring up new nodes, please note that you will need to complete a joining request at http://www.planet-lab.org/joining. Thanks in anticipation. If you have any queries, please do not hesitate to contact me. Paul Brett PlanetLab Support Email: pau...@pl... Tel No: +1 503 712 4520 |
|
From: Paul B. <pau...@pl...> - 2003-11-10 17:52:16
|
Massimo | -----Original Message----- | From: pla...@li... | [mailto:pla...@li...] On | Behalf Of Massimo Canonico | Sent: Monday, November 10, 2003 7:22 AM | To: pla...@li... | Subject: [Planetlab-users] [Fwd: Two questions] | | Hi, | I'm new in PlanetLab and I would ask two questions about | iteration | between machines and software installation: | | | 1) To compare the performance of my application, I need to | install some | software in the nodes. This software needs to be installed | as root. Is | it possible install this software as root? Can someone do | it for me | (i.e., PlanetLab administrators)? Within your vserver account, you can su to root and install any software you require. | | 2) my application monitoring different PCs around internet | measuring the | bandwidth between any machine in a group. To evaluate this | bandwidth I | have implemented a Client/Server application. Is there a firewall | setting between the nodes in PlanetLab? PlanetLab nodes should not be firewalled from the internet. If you find an exception, please let us know and we will work to fix the issue. If you have any queries, please do not hesitate to contact me. Paul Brett PlanetLab Support Email: pau...@pl... Tel No: +1 503 712 4520 |
|
From: Adams, R. <rob...@in...> - 2003-11-10 17:50:28
|
We don't enable root access to the system proper. But, each vserver has a 'virtual root' account that allows the installing of RPMs and other system modifications. Log into your vserver with SSH and then 'su' your vserver's virtual root account. For networking, there are 'safe raw sockets' (http://www.planet-lab.org/raw_sockets) which make many raw socket operations available. As to firewalls, in general, PlanetLab nodes are supposed to=20 be on the open Internet and not have any firewall between them and the network. But the realities of site administration and the wild-west Internet has led to some port blocking on some of the nodes. There is not a list of these. -- RA -----Original Message----- From: pla...@li... [mailto:pla...@li...] On Behalf Of Massimo Canonico Sent: Monday, November 10, 2003 7:22 AM To: pla...@li... Subject: [Planetlab-users] [Fwd: Two questions] Hi, I'm new in PlanetLab and I would ask two questions about iteration=20 between machines and software installation: 1) To compare the performance of my application, I need to install some=20 software in the nodes. This software needs to be installed as root. Is=20 it possible install this software as root? Can someone do it for me=20 (i.e., PlanetLab administrators)? 2) my application monitoring different PCs around internet measuring the bandwidth between any machine in a group. To evaluate this bandwidth I=20 have implemented a Client/Server application. Is there a firewall=20 setting between the nodes in PlanetLab? Thanks, Massimo. ------------------------------------------------------- This SF.Net email sponsored by: ApacheCon 2003, 16-19 November in Las Vegas. Learn firsthand the latest developments in Apache, PHP, Perl, XML, Java, MySQL, WebDAV, and more! http://www.apachecon.com/ _______________________________________________ Planetlab-users mailing list Pla...@li... https://lists.sourceforge.net/lists/listinfo/planetlab-users |
|
From: Massimo C. <can...@mf...> - 2003-11-10 15:20:42
|
Hi, I'm new in PlanetLab and I would ask two questions about iteration between machines and software installation: 1) To compare the performance of my application, I need to install some software in the nodes. This software needs to be installed as root. Is it possible install this software as root? Can someone do it for me (i.e., PlanetLab administrators)? 2) my application monitoring different PCs around internet measuring the bandwidth between any machine in a group. To evaluate this bandwidth I have implemented a Client/Server application. Is there a firewall setting between the nodes in PlanetLab? Thanks, Massimo. |
|
From: Yun M. <ma...@ci...> - 2003-11-04 23:01:36
|
Hi, I got a couple of questions: 1. Suppose I screw up my files in the vserver, how to "format" the fs to get back from the beginning? or I have to bug the planetlab support to do this for me? 2. How to start my service each time when the boxes are rebooted? The /etc/rc.d/rc.local seems to be the answer (as usual), but I can't find related docs. 3. Suppose I upload my code to one particular node, do I have to write a script to do this for all nodes, or there is already some handy software that I can sync the rest of nodes to that one? It seems to me that this is not as trivial as I originally thought because the configuration for different nodes may be different, and in particular, sometimes the fs may need to be considered as soft state. ;) Thanks in advance. Yun |
|
From: Timothy R. <tr...@in...> - 2003-11-04 22:22:44
|
Tomasz, Assuming that you've uploaded your ssh public key to the PlanetLab web site, the next step is to get a slice. Berlin has a number of slices, so what you need to do is to get your PI (Principle Investigator) to authorize you to use one or more slices. Once this is done, you can just ssh to any planetlab node as the name of your slice, using your private key. For instance, I'm authorized to use slice "irb9" here at the Lab, so I do: $ ssh -i ~/.ssh/planetlab_rsa ir...@ed... [irb9@edi irb9]$ Good luck! -- Mothy On Tue, 2003-11-04 at 13:14, tomas wrote: > Hi! > > I am new to planet-lab. I have allready an account and i have allready > read the paper about raw-sockets in planet-lab. > Is there a paper about how to start my service on planet lab? First i > only want to start a foo-service > (a service which do nothing) on a node. > > Thanks a lot for helping > > Regards > Tomasz Gingold > tkn-berlin |
|
From: tomas <tgi...@gm...> - 2003-11-04 21:14:29
|
Hi! I am new to planet-lab. I have allready an account and i have allready = read the paper about raw-sockets in planet-lab. Is there a paper about how to start my service on planet lab? First i = only want to start a foo-service (a service which do nothing) on a node. Thanks a lot for helping Regards Tomasz Gingold tkn-berlin |
|
From: Minseok K. <kw...@cs...> - 2003-10-30 20:32:58
|
Thanks. It works now. On Thu, 30 Oct 2003, Neil Spring wrote: > > On Oct 30, 2003, at 11:09 AM, Bowman, Mic wrote: > > > The passphrase is a property of your key. The man pages for ssh > > describe > > how to change or manipulate the passphrases on keys. > > and the manpage of ssh-agent describes how to cache decrypted private > keys so that you only have to type your passphrase once. > > passphrase-less keys are bad. > > -neil > > > > > --Mic > > > > -----Original Message----- > > From: pla...@li... > > [mailto:pla...@li...] On Behalf Of > > Minseok Kwon > > Sent: Thursday, October 30, 2003 09:48 AM > > To: pla...@li... > > Subject: [Planetlab-support] SSH on planetlab > > > > > > Hi, > > > > Is there a way to get around the passphrase in SSH on planetlab? I > > tried > > .rhosts but doesn't work. Thanks. > > > > Best regards, > > > > M. S. Kwon > |
|
From: Neil S. <ns...@cs...> - 2003-10-30 19:31:31
|
On Oct 30, 2003, at 11:09 AM, Bowman, Mic wrote: > The passphrase is a property of your key. The man pages for ssh > describe > how to change or manipulate the passphrases on keys. and the manpage of ssh-agent describes how to cache decrypted private keys so that you only have to type your passphrase once. passphrase-less keys are bad. -neil > > --Mic > > -----Original Message----- > From: pla...@li... > [mailto:pla...@li...] On Behalf Of > Minseok Kwon > Sent: Thursday, October 30, 2003 09:48 AM > To: pla...@li... > Subject: [Planetlab-support] SSH on planetlab > > > Hi, > > Is there a way to get around the passphrase in SSH on planetlab? I > tried > .rhosts but doesn't work. Thanks. > > Best regards, > > M. S. Kwon |
|
From: Bowman, M. <mic...@in...> - 2003-10-30 19:09:28
|
The passphrase is a property of your key. The man pages for ssh describe how to change or manipulate the passphrases on keys. --Mic -----Original Message----- From: pla...@li... [mailto:pla...@li...] On Behalf Of Minseok Kwon Sent: Thursday, October 30, 2003 09:48 AM To: pla...@li... Subject: [Planetlab-support] SSH on planetlab Hi, Is there a way to get around the passphrase in SSH on planetlab? I tried .rhosts but doesn't work. Thanks. Best regards, M. S. Kwon ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ Planetlab-support mailing list Pla...@li... https://lists.sourceforge.net/lists/listinfo/planetlab-support |
|
From: Adams, R. <rob...@in...> - 2003-10-30 00:35:51
|
Because of problems with access to nodes because of overfull disks,=20 the following PlanetLab nodes will be reinstalled on Thursday October 30th: pl1.ece.toronto.edu pl2.ece.toronto.edu s2_803.ie.cuhk.edu.hk If you had experiments or services running on these PlanetLab nodes, you will need to reinstall. Future plans are to implement disk quotas and dynamic slices -- both of which will reduce this problem. But, in the short term, we are sorry for this inconvience. -- Robert Adams |
|
From: Michal S. <mi...@cs...> - 2003-10-29 15:40:07
|
Hi Andy, > I can't reproduce your problem -- I see both incoming and outgoing messages > for a connection, including the SYNACK. I've attached three files which may > help you: > [..] Thanks! It works indeed. After you convinced me with your piece of code that it is not the PlanetLab's limitation, I sat down and found one stupid bug in my own code.. (I was filtering out all packets that do not target MYPORTNUMBER -- obviously, the outgoing packets don't :-)) Sorry for bothering you, and thanks once again for your help :-) Kind regards, M. -- Michal Szymaniak | mailto:mi...@cs... | http://www.cs.vu.nl/~mszyman |
|
From: Andy B. <ac...@CS...> - 2003-10-29 13:32:47
|
Hi Michal, I can't reproduce your problem -- I see both incoming and outgoing messages for a connection, including the SYNACK. I've attached three files which may help you: sniffdump.c: Opens a safe raw socket on a particular port, attaches a pcap header to read packets, and prints them to stdout. plabdump: Invokes sniffdump and then pipes its output to tcpdump for nice formatting. tcp_port.c: Creates a connection to www.cs.princeton.edu from local port 12345 Note that sniffdump and plabdump are also available as part of the plkmodutil package: http://www.planet-lab.org/raw_sockets/notes_files.html Here is some output from planetlab-2.cs.princeton.edu. I see all packets sent between the web server and the local host. [princeton8@planetlab-2 testit]$ ./plabdump -l tcp port 12345 13:05:45.663254 planetlab-2.cs.princeton.edu.12345 > web0.CS.Princeton.EDU.http: S 3930150656:3930150656(0) win 5840 <mss 1460,sackOK,timestamp 309102267 0,nop,wscale 0> (DF) 13:05:45.664677 web0.CS.Princeton.EDU.http > planetlab-2.cs.princeton.edu.12345: S 2193066063:2193066063(0) ack 3930150657 win 24616 <nop,nop,timestamp 9227789 309102267,nop,wscale 0,nop,nop,sackOK,mss 1460> (DF) 13:05:45.664731 planetlab-2.cs.princeton.edu.12345 > web0.CS.Princeton.EDU.http: . ack 1 win 5840 <nop,nop,timestamp 309102268 9227789> (DF) 13:05:46.673981 planetlab-2.cs.princeton.edu.12345 > web0.CS.Princeton.EDU.http: F 1:1(0) ack 1 win 5840 <nop,nop,timestamp 309102369 9227789> (DF) 13:05:46.674631 web0.CS.Princeton.EDU.http > planetlab-2.cs.princeton.edu.12345: . ack 2 win 24616 <nop,nop,timestamp 9227890 309102369> (DF) 13:05:46.675078 web0.CS.Princeton.EDU.http > planetlab-2.cs.princeton.edu.12345: F 1:1(0) ack 2 win 24616 <nop,nop,timestamp 9227890 309102369> (DF) 13:05:46.675106 planetlab-2.cs.princeton.edu.12345 > web0.CS.Princeton.EDU.http: . ack 2 win 5840 <nop,nop,timestamp 309102369 9227890> (DF) Cheers, Andy ----- Original Message ----- From: "Michal Szymaniak" <mi...@cs...> To: "Andy Bavier" <ac...@CS...> Cc: <pla...@li...> Sent: Tuesday, October 28, 2003 3:29 AM Subject: Re: a question about raw sockets > > Andy, > > > To monitor incoming and outgoing packets on a port, PlanetLab provides raw > > "sniffer" sockets. Here is a bit of documentation: > > > > http://www.planet-lab.org/raw_sockets/api_sniffer.html > > > > I followed the instructions exactly, and: > - I can see incoming raw IP packets :-) > - I still cannot see any outgoing ones :-( > For example, during a TCP-handshake, I can see incoming SYN and ACK > packets, but I cannot see any SYNACK packet between them. > > Here is a piece of my code: > > #define PACKET_SIZE 1500 > > int m,on; > char packet[PACKET_SIZE]; > > if ((m=socket(PF_INET,SOCK_RAW,IPPROTO_TCP))==-1) > fatal("monitor:socket"); > on=1;if (setsockopt(m,SOL_SOCKET,SO_REUSEADDR,&on,sizeof(int))==-1) > fatal("monitor:reuseaddr"); > on=1;if (setsockopt(m,0,SO_RAW_SNIFF,&on,sizeof(int))==-1) > fatal("monitor:setrawsniff"); > if (bind(m,(struct sockaddr*)&sa,sizeof(struct sockaddr_in))==-1) > fatal("monitor:bind"); > > and then loop on recvfrom(m,packet,PACKET_SIZE,0,NULL,NULL); > > > Any ideas what may be wrong? BTW, thanks to everyone who tried to help me > the last time :-) > > Kind regards, > M. > -- > Michal Szymaniak | mailto:mi...@cs... | http://www.cs.vu.nl/~mszyman > |
|
From: Rob <ca...@cs...> - 2003-10-28 17:40:21
|
On Tue, Oct 28, 2003 at 09:42:39AM +0100, Michal Szymaniak wrote: > > Andy, > > > To monitor incoming and outgoing packets on a port, PlanetLab provides raw > > "sniffer" sockets. Here is a bit of documentation: > > > > http://www.planet-lab.org/raw_sockets/api_sniffer.html > > > > I followed the instructions exactly, and: > - I can see incoming raw IP packets :-) > - I still cannot see any outgoing ones :-( > For example, during a TCP-handshake, I can see incoming SYN and ACK > packets, but I cannot see any SYNACK packet between them. I am fairly sure this is a limitation of raw sockets and the linux netfilter kernel architechture (see http://www.netfilter.org/ for details) - basically out going packets don't get looked at with SOCK_RAW. Unfortunately, I don't know of a way to get around this in the planetlab environment... normally I would use libpcap to get around this, but we don't have permissions: [umd2@planet-lab nload]$ ./nload -i eth0 Device: eth0 ER: pcap_open_live: socket: Operation not permitted Does anyone else know a way around this? - Rob . |
|
From: Michal S. <mi...@cs...> - 2003-10-28 08:43:25
|
Andy, > To monitor incoming and outgoing packets on a port, PlanetLab provides raw > "sniffer" sockets. Here is a bit of documentation: > > http://www.planet-lab.org/raw_sockets/api_sniffer.html > I followed the instructions exactly, and: - I can see incoming raw IP packets :-) - I still cannot see any outgoing ones :-( For example, during a TCP-handshake, I can see incoming SYN and ACK packets, but I cannot see any SYNACK packet between them. Here is a piece of my code: #define PACKET_SIZE 1500 int m,on; char packet[PACKET_SIZE]; if ((m=socket(PF_INET,SOCK_RAW,IPPROTO_TCP))==-1) fatal("monitor:socket"); on=1;if (setsockopt(m,SOL_SOCKET,SO_REUSEADDR,&on,sizeof(int))==-1) fatal("monitor:reuseaddr"); on=1;if (setsockopt(m,0,SO_RAW_SNIFF,&on,sizeof(int))==-1) fatal("monitor:setrawsniff"); if (bind(m,(struct sockaddr*)&sa,sizeof(struct sockaddr_in))==-1) fatal("monitor:bind"); and then loop on recvfrom(m,packet,PACKET_SIZE,0,NULL,NULL); Any ideas what may be wrong? BTW, thanks to everyone who tried to help me the last time :-) Kind regards, M. -- Michal Szymaniak | mailto:mi...@cs... | http://www.cs.vu.nl/~mszyman |
|
From: Larry P. <ll...@CS...> - 2003-10-23 17:45:01
|
As you've heard me say before, it would be very helpful to have PlanetLab sites provide a local BGP feed. To help this process along, I've drafted a note (see attached) that you can edit and send to your local network administrator. (The message explains why the feed would be valuable, in case you need to be convinced before asking your admin folks.) Note that your dept router may not give us what we really need; it's most likely the case that we need a feed from your campus BGP router. Larry |
|
From: Kiyohide N. <nak...@cr...> - 2003-10-20 02:57:44
|
PlanetLab users: planetlab1.koganei.wide.ad.jp will be disconnected at around 23:00 PM, GMT, on Oct 20th. At the same time, IP address of the node wil be changed from 202.249.37.220 to 202.249.37.211. I'm sorry for your any inconvenience. Regards Kiyohide Nakauchi ************************************************************* Kiyohide Nakauchi, Ph.D. Researcher Internet Architecture Group Information and Network Systems Division Communications Research Laboratory Incorporated Administrative Agency 4-2-1, Nukui-kitamachi, Koganei, Tokyo, 184-8795, Japan Phone : +81-42-327-5403 FAX : +81-42-327-6680 Email : nak...@cr... URL : http://www.crl.go.jp/jt/a113/member/nakauchi/ ************************************************************* |
|
From: Zhen W. <zw...@cs...> - 2003-10-16 20:27:26
|
Hi, everyone:
This afternoon (Thursday - Oct 16, 2003), there will a power down for the
planetlab machines in our school. Please pay attention to it, especially
when your work are running on it!
Sorry for the inconvenience!
****************************************
Name: Zhen Wu
Email: zw...@cs...
Phone: 541-346-0199
Address: CIS Dept, University Of Oregon
Eugene, OR 97403
****************************************
|
|
From: Andy B. <ac...@CS...> - 2003-10-10 14:06:57
|
I forgot to put the ping source that has been modified for safe raw sockets into the SourceForge CVS repository. In the meantime you can get it here: http://www.cs.princeton.edu/~acb/iputils.tgz Grep for "#ifdef PLANETLAB" in the source; there are only a couple of changes. Andy |
|
From: Akihiro N. <na...@cs...> - 2003-10-09 13:55:02
|
If you have not looked at this, take a look. http://www.planet-lab.org/raw_sockets/ --Aki -----Original Message----- From: pla...@li... [mailto:pla...@li...] On Behalf Of TIAN Rui-xiong Sent: Thursday, October 09, 2003 9:01 AM To: pla...@li... Subject: [Planetlab-users] problem about raw socket in PlanetLab I need your help on planetlab raw socket APIs. I wrote a ping using the safe raw socket, but it does not work.=3D In my program, once I wrote bytes to a created raw socket, a=3D parameter error always occured. The codes of my ping are= nearly=3D the same as those in Richard Steven's "UNIX Networking=3D Programming" except those which bind the socket to an ICMP=3D identifier. If I compile it on a normal linux machine, the=3D program works correctly. Maybe you can give me the source code of planet-lab ping or=3D some advice. Thank you very much! Yours, Ruixiong Tian ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. SourceForge.net hosts over 70,000 Open Source Projects. See the people who have HELPED US provide better services: Click here: http://sourceforge.net/supporters.php _______________________________________________ Planetlab-users mailing list Pla...@li... https://lists.sourceforge.net/lists/listinfo/planetlab-users |
|
From: TIAN Rui-x. <rx...@ns...> - 2003-10-09 13:04:53
|
|
From: TIAN Rui-x. <rx...@ns...> - 2003-10-09 13:00:25
|
I need your help on planetlab raw socket APIs.
I wrote a ping using the safe raw socket, but it does not work.=3D
In my program, once I wrote bytes to a created raw socket, a=3D
parameter error always occured. The codes of my ping are=
nearly=3D
the same as those in Richard Steven's "UNIX Networking=3D
Programming" except those which bind the socket to an ICMP=3D
identifier. If I compile it on a normal linux machine, the=3D
program works correctly.
Maybe you can give me the source code of planet-lab ping or=3D
some advice. Thank you very much!
Yours,
Ruixiong Tian
|
|
From: Sedayao, J. <jef...@in...> - 2003-10-09 01:43:36
|
PlanetLab Users: We have found and fixed a problem with disappearing vservers. All = slices created after a PlanetLab host was initially set up would be be = deleted. We have believe that we fixed the problem, but please let us = know if the problem reoccurs. Thanks. Jeff Sedayao PlanetLab Support |
|
From: Patrick G. B. <br...@cs...> - 2003-10-08 19:32:15
|
pl1.cs.unm.edu and pl2.cs.unm.edu will go down tomorrow (Oct. 8) at 10:00am MDT to install a network addressable power supply. They should be back up shortly thereafter. -- Patrick G. Bridges br...@cs... GPG ID = CB074C71 GPG fingerprint = FEEA ECFF 1E23 148C 2804 FDD9 DB63 6993 CB07 4C71 "Anyone that can't make money on Sports Night should get out of the money-making business" - Calvin, on the last episode of Sports Night |
