Hello,
for about 2 weeks I've been noticing (I think) problems with one of my client's Edge Gateway. Remote IPs sometimes (randomly) cant connect to the services which are on the VM behind the edge ( to the VM with CPANEL installed which is behind the NAT) on random ports (generally 80,143,110,993,995, some cpanel ports),
In client Edge Gateway log files I see tousands of this (many , different IP addresses)
2017-06-19T01:50:31+00:00 vse-1031e7fc-0450-44c8-8a1a-851f0c2afc15-0 kernel[]: [126884ec-bbd2-4738-973f-d3f9df6ad6fd]: [kern.info] nf_conntrack: TCP Connection Timedout 32.88.36.9:17200 -> 181.21.22.147:80
2017-06-19T01:50:31+00:00 vse-1031e7fc-0450-44c8-8a1a-851f0c2afc15-0 kernel[]: [126884ec-bbd2-4738-973f-d3f9df6ad6fd]: [kern.info] nf_conntrack: TCP Connection Timedout 32.88.36.9:17201 -> 181.21.22.147:80
2017-06-19T01:51:48+00:00 vse-1031e7fc-0450-44c8-8a1a-851f0c2afc15-0 kernel[]: [126884ec-bbd2-4738-973f-d3f9df6ad6fd]: [kern.info] nf_conntrack: TCP Connection Timedout 81.23.115.158:43135 -> 181.21.22.147:80
In one minute Edge logs many of above.
Edge was changed from Compact size to X-Large size about 3 weeks ago. Edge version 6.2.4. There is no firewall in vcloud (in vsphere is enabled by default but accept any any). There are some DNATs and SNATs
Our infrastructure looks:
Clients Edges are connected to the main EDGE which is connected later to devices connected to the internet.
What is interesting - we started some scripts which were making hundred of connections and for example any of these connections were timed out any of there were logged in log files.
Anybody have some ideas? I will be really appreciate for any help
UP!
Anybody have some idea?
Haven't seen this error before . Looking at the snippet TCP connections are timing-out . First and foremost i would rule out if it is an Host specific issue ,so worth check Uplinks and do a test on another host by migrating the edge. May be some TCP tuning might be required if it not a Edge specific issue.