• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 476
  • Last Modified:

rcp problems

I have an automated process using rcp between Sun and Alpha systems. Sometimes I get an error of "rcp failure connection lost".

Can anyone suggest why?
0
WakiBaki
Asked:
WakiBaki
  • 5
  • 4
1 Solution
 
mapcCommented:
Could you provide more details, file size, how often it happends, when this task is run, etc.
0
 
salikovaCommented:
you would test line between computers, for example
ping -s name_host
 please see how many packets is lost
0
 
WakiBakiAuthor Commented:
The files are quite small, typically no more than 10/15k. So far this has happened on 4 occasions over about 1 month, and these have all been evening / overnight. The frequency and times of the actions are variable, as the process is initiated elsewhere (though our ecom offering).
0
Concerto's Cloud Advisory Services

Want to avoid the missteps to gaining all the benefits of the cloud? Learn more about the different assessment options from our Cloud Advisory team.

 
mapcCommented:
Do you see any messages in kernel.
It's possible that under extremely heavy load the machine will drop packets.
salikova's answer is still not an answer, just one step to it :)
0
 
WakiBakiAuthor Commented:
The connection appears to be sound and has lost no packets when carrying out a ping(s). According to our switch (Cisco) there is no packet loss of any consequence (2 packets in 3 months on one system, 3 on the other. the latter having about 1200 concurrent users).
0
 
mapcCommented:
It could happen if the process (in.rshd) dies or was killed.
Does it make sense in your environment?
Any messages in syslog?
0
 
WakiBakiAuthor Commented:
The connection appears to be sound and has lost no packets when carrying out a ping(s). According to our switch (Cisco) there is no packet loss of any consequence (2 packets in 3 months on one system, 3 on the other. the latter having about 1200 concurrent users).
0
 
WakiBakiAuthor Commented:
This might lead on to something...

If the rcp made a successful connection and the process was immediately killed on the destination host (before completion) could this produce the error?
0
 
mapcCommented:
The answer is yes.
In that case the filedescriptor is closed by the underlying OS and this error is produced at the client side.

I *think* some OSs implemented "kill _user_ processes when low on resources" algorithms, don't know if it's still true. e.g. user: uid<=someuid
0
 
WakiBakiAuthor Commented:
This makes sense, but i will need to do some further investigation to prove it 100%. I am happy that this is the most likely cause.

Thanks for the help!
0

Featured Post

Concerto's Cloud Advisory Services

Want to avoid the missteps to gaining all the benefits of the cloud? Learn more about the different assessment options from our Cloud Advisory team.

  • 5
  • 4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now