0

I have come across a problem with SSH and Telnet wherein the connectivity is slow and drops multiple times. But I see that my network has no problems. When I ping a server, it gives me ttl of 127, packet losses 0% and time is less than 1 ms. This is a centOS 6.9 2.6.32-696.20.1 server. Its an app server running tomcat services. The initial problem was that this app server looses connectivity while communicating with the DB server (MSSQL 2012). I am facing this issue from almost 3 days, but did not find anything that triggers it. I have not found any error events in /var/log/messages. I have not found any hardware issues on the server. In ifconfig the errors:0 dropped:0

I ran the below command to test telnet connectivity from my app server:

while true; do sleep 1 | telnet xx.xx.xx.xx 1433 ; done;   //1433 is the allowed port for DB

I get the message : Escape Character is 'CTRL+]' But I see that sometimes it tries to connect, it takes a long time and then it says : Connection refused I do not see any pattern to this "Connection Refused" message, its just a random occurrence. I have stopped all the tomcat6 services, stopped iptables to test it, and still see no difference.

While connecting the server through ssh, I see that it takes a lot of time to get me logged in, and some time it says "Connection Refused"

I have also done a tcp dump on the network port and on the destination host, but dont see any packet losses.

I have compared it with the other servers that are of the same specs(including BIOS and Hardware), same kernel version, same packages and the in the same chassis (because I wanted to check whether or not the MXL switch is causing any issues). I have not seen any issues with regards to SSH and Telnet on them and there is no issues with the DB connectivity on them.

Does any one have any troubleshooting tips that I can try? or is there some fix for it that I'm unaware of?

Telnet version: telnet-0.17-48.el6.x86_64 SSH version: openssh-5.3p1-123.el6_9.x86_64 Server: Dell M820, iDrac firmware and life cycle controller version: 2.30.30.30, Bios version: 1.5.1 Note I have not done any changes with the Telnet or SSH settings.

  • 1
    can you check you don't have two systems using the same IP on the server's side? this kind of symptom can be seen with two servers fighting over the same IP (and if one server is mostly idle, it won't be seen much). So you could try changing the IP to a known free one and compare. Of course there are dozens of possible other causes – A.B May 1 '18 at 13:48
  • For some reason after I failed over the SQL DB cluster to the other node. I do not see such issue. – dannydsouza0908 May 21 '18 at 7:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.