vBulletin Search Engine Optimization
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| BismillahirRehmanirRahim Hi Admins, We are facing very strange problem when copying files to one of our Server. when we do scp any file [ size e.g.20MB] to this server the scp stuck with "-stalled-" status. while for very small files in KB's it work. i tried rsync and ftp as well. they also faile to copy more than 100KB. there is no connectivity problem. we are able to connect / ftp / telnet / ssh to this server. but facing this problem in copying files. The servers was working fine before we moved it to new location [new IP , new Subnet]. I am attaching the test log, if you need any more information then plz email me at msfaheem at yahoo.com Appreciate your prompt feedback. Your Junior Admin Salman -------------------------------------------------------------------------------------------------------------------------------------- $ scp Arch.Z prod1:/data/ Arch.Z 1% | | 144 KB - stalled - ^C Killed by signal 2. $ $ date > test_date $ scp test_date dcpcore1:/data/ test_date 100% | ************************************************** *******************************| 29 00:00 $ # ifconfig -a lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1 inet 127.0.0.1 netmask ff000000 eri0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2 inet 10.44.10.1 netmask ffffff80 broadcast 10.44.10.127 ether 0:14:4f:1e:5c:5 -------------------------------------------------------------------------------------------------------------------------------------- |
| |||
| On Aug 30, 4:37 am, [email protected] wrote: > BismillahirRehmanirRahim > > Hi Admins, > > We are facing very strange problem when copying files to one of our > Server. > > when we do scp any file [ size e.g.20MB] to this server the scp stuck > with "-stalled-" status. > > while for very small files in KB's it work. > > i tried rsync and ftp as well. they also faile to copy more than > 100KB. > > there is no connectivity problem. we are able to connect / ftp / > telnet / ssh to this server. > > but facing this problem in copying files. > > The servers was working fine before we moved it to new location [new > IP , new Subnet]. > > I am attaching the test log, if you need any more information then plz > email me at msfaheem at yahoo.com > > Appreciate your prompt feedback. > > Your Junior Admin > Salman > > ---------------------------------------------------------------------------*----------------------------------------------------------- > > $ scp Arch.Z prod1:/data/ > Arch.Z 1% > | > | 144 KB - stalled - > ^C > Killed by signal 2. > > $ > > $ date > test_date > $ scp test_date dcpcore1:/data/ > test_date 100% | > ************************************************** ********************************| > 29 00:00 > $ > > # ifconfig -a > lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index > 1 > inet 127.0.0.1 netmask ff000000 > eri0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 > index 2 > inet 10.44.10.1 netmask ffffff80 broadcast 10.44.10.127 > ether 0:14:4f:1e:5c:5 > > ---------------------------------------------------------------------------*----------------------------------------------------------- force both interfaces to full duplex and see what happens |
| |||
| John Schmidt <[email protected]> writes: >omie wrote: >> force both interfaces to full duplex and see what happens >That's an *astoundingly* bad idea for any gear manufactured in the >last 10 years. Besides, a duplex mismatch will cause extreme slowness, not a permanent stall. I think it's a middle box which just doesn't properly maintain connection state or which doesn't deal with window sizes properly. Casper -- Expressed in this posting are my opinions. They are in no way related to opinions held by my employer, Sun Microsystems. Statements on Sun products included here are not gospel and may be fiction rather than truth. |
| |||
| Casper H.S. Dik <[email protected]> wrote: > John Schmidt <[email protected]> writes: > >>omie wrote: > >>> force both interfaces to full duplex and see what happens > >>That's an *astoundingly* bad idea for any gear manufactured in the >>last 10 years. > > Besides, a duplex mismatch will cause extreme slowness, not a permanent > stall. > > I think it's a middle box which just doesn't properly maintain > connection state or which doesn't deal with window sizes properly. > > Casper I would agree with Casper on that. Traceroute will tell you the intermediate steps along the way. Perhaps you could attempt the file transfer to and from some intermediate locations. Can you try the transfers between two different machines on the same subnets as these ones, to be certain that it is the network and not the machines themselves? -- Dr Tristram J. Scott Energy Consultant |
| |||
| In article <[email protected] .com>, [email protected] wrote: > there is no connectivity problem. we are able to connect / ftp / > telnet / ssh to this server. > > but facing this problem in copying files. I've seen this with MTU mismatches, typically when something along the way is blocking ICMP type 3 packets. Try pinging with different size packets until it fails, then set your MTU to that less the 40-byte TCP hdr. -- Sak Wathanasin Network Analysis Limited http://www.network-analysis.ltd.uk |
| ||||
| On 31 Ago, 13:06, [email protected] (Tristram Scott) wrote: > Casper H.S. Dik <[email protected]> wrote: > > > > > > > John Schmidt <[email protected]> writes: > > >>omie wrote: > > >>> force both interfaces to full duplex and see what happens > > >>That's an *astoundingly* bad idea for any gear manufactured in the > >>last 10 years. > > > Besides, a duplex mismatch will cause extreme slowness, not a permanent > > stall. > > > I think it's a middle box which just doesn't properly maintain > > connection state or which doesn't deal with window sizes properly. > > > Casper > > I would agree with Casper on that. > > Traceroute will tell you the intermediate steps along the way. Perhaps you > could attempt the file transfer to and from some intermediate locations. > > Can you try the transfers between two different machines on the same > subnets as these ones, to be certain that it is the network and not the > machines themselves? > > -- > Dr Tristram J. Scott > Energy Consultant - Nascondi testo tra virgolette - > > - Mostra testo tra virgolette - You can debug your ssh transfer file through truss command. You must retrieve the pid of the ssh process and you can launch truss - a -f -p <pid_of_ssh_process> for watching in real time what happens! Alessio |