AWS Server - Can't get vps token #3737
Comments
|
Check in |
|
It is correct: But today no outage happened. But the VPS journal output of omr-admin has all the entries. So it seems another issue not related to my issue. |
|
Check also in /etc/shorewall6/params.net |
|
Ahhh there it is: I will watch now. But your testing feedback with AWS would be very welcome, too. Maybe i overlook something specific there. |
|
I had no outage for 1,5 days. But did a speedtest right now and the issue happens again. The server does not respond anymore. I guess it is something on the AWS side. Some restriction i don't know. Would be great to find the cause there. |
|
I have tested the same setup now but with the VPS on Hetzner.de. No issue so far. Slightly lower aggr speed but stable. No speedtest is resulting in a VPS outage. I assume it is something specific on the AWS side. By the way after the VPS changes on the omr side i can't use iperf anymore: i understand i shouldn't install iperf from the openwrt repo. But how can i reinstall it? What means segmentation fault? And my PPPOE uplink is always marked as down but it is not: |
|
iperf3 is already available, the omr-iperf script only add the key needed for VPS iperf server. |
But i can't use the omr-iperf script because of "Segmentation fault". How can i fix that? The pppoe output looks good but still marked as down. |
|
Might be of some help. Ive been using an AWS lightsail VPS for the past 2-3 weeks. I have not had any issues like this at all. I'm not using any snapshots though just the stable version of both server and router. I also get the segmentation fault though trying to use iperf. Messed about with for a bit but gave up. It didnt crash my VPS though. I've moved back to my main VPS now on ionos. The reason I moved to aws is I had a similar but not that same issue on ionos where one day the vps just stopped working. I hadn't made any changes. My backup VPS also stopped working aws was the only one that was working. A week later the ionos server started working again. I had made no changes. I mention this because as you've highlighted yourself don't rule out a vps provider issue. |
|
@dersch81 iperf may crash due to wrong key, I don't know. You can restore keys with System->OpenMPTCProuter, Wizard tab, "advanced settings" checkbox, and "Force retrieve settings" checkbox. For the pppoe device I don't see any OMR-Tracker log, this seems strange. Can you restart it via |
|
omr-iperf is working now. Did the force retrieve (which i did several times before) again. But also not correct: after restarting the omr-tracking service i just see this: The tracker itself is working: 
As you can see there is also a same VDSL line via DHCP on the same POP connected DTAG-DIAL24. Just my local PPPOE is not tracked but working in aggr mode. |
|
I would need |
It is referencing the wrong WAN |
|
Strange, I would need |
|
|
A new snapshot image is compiling, I hope this will fix the issue... |
|
Hi @Ysurac is the snapshot done? |
|
Yes., available on https://snapshots.openmptcprouter.com/6.6/ |
|
Ok installed v0.62 Snapshot from 8th Januar The symptom has changed but not resoved: 
That's correct now: I tried to restart the tracker manually: (eth2 is really down and another WAN) |
|
the service_data error, is due to an OpenWRT change, not a big problem. |
|
Yes that's ok. But the PPPOE Tracker issue is still there. But changed from red to orange. So close to green :D |
|
Do you have a message with the orange ? Orange is a warning only. |
|
Just "some connectivity tests failed" But i have no clue what kind of test. 
|
Expected Behavior
Stable connection from Router to VPS
Current Behavior
The connection, aggr Speed and so on is great with AWS. But it just runs for a very short time. Then this happens:
Router Syslog:
VPS
journalctl -u omr-adminThe AWS interface is ens5
It never recovers. Only rebooting the AWS VPC instance is fixing it for a time.
The AWS VPC is not reachable anymore. The issue starts right after benching it with speedtests, or just after a while connected.
I already found #3733 - the issue is not the same but i installed the latest dev branch you mentioned. Did not fix it.
Besides i have on the router side this all the time:
Possible Solution
No idea
Steps to Reproduce the Problem
Context (Environment)
I try to have a stable mptcp setup. Had a lot of instability with HT-Hosting.de. Decided to switch to AWS. It is much better when it works. More Down and Up bandwith in aggr mode. But now not useable at all in general.
Specifications
The text was updated successfully, but these errors were encountered: