Is there a way to save the log files for review after a crash / reboot? Possibly to a external drive used for system backup. I have a server pure.
First off, when this happens, check start.local or the server’s IP address in your browser to see if it’s in diagnostic mode. If you still get nothing, try SSH’ing in and run:
sudo journalctl -xefa
You can also open a tab in the browser with the StartOS logs and check them after a crash. There will be a download button at the bottom to easily save them in HTML format. Make sure to do this before rebooting the server—you might gather more information about what’s going on.
The servers are glitching and freezing. This happens unexpectedly days or weeks between issues. Power light on server on but cannot ssh or use monitor and keyboard direct. A power cycle reboots and it starts and no problems. But the logs are restarted after the reboot. I’m wanting to see pre-crash logs. I already replaced the small power supplies. I have 2 servers, different apps, same problem.
The best option would be to keep the log tabs open in the browser and see if this catches any errors after a crash.
Are you saying you have two servers that are showing the same symptoms? Could be a power supply issue?
There is no log file showing events before a crash. New log files seem to be created on startup.
I have two servers and had to get you to send me the correct power supplies. But the issues only decreased with the larger power supplies you sent me.
I use one for Bitcoin related apps, the other I mess around with the other apps that I find interesting.
I ran start9 on my own servers, one was an old Thinkbox and the other was a Mac mini. They were very reliable and the same apps. I bought the servers from you to support your great buisness.
I want to see pre-crash logs, how do I record those permanently and continuously
Hi Dtex,
I’m looking into the issues you’ve been experiencing with freezing and glitching. I have a few questions to help narrow things down:
-
When the server freezes, can you check if it can still be pinged and if it shows as connected to your router?
-
To ping the server:
- On your computer, open a terminal (or Command Prompt in Windows).
- Type
ping [server IP address]and hit Enter. - If the server is reachable, you’ll see responses like “Reply from [server IP address].”
- If it’s not reachable, you’ll see something like “Request timed out.”
-
To check if the server is connected to your router:
- Log into your router’s web interface (you can usually access it by entering your router’s IP address in a web browser).
- Look for a section called “Connected Devices,” “Device List,” or something similar.
- See if the server’s IP address or hostname is listed there. If it’s shown as connected, that’s a good sign the server is still up, even if it’s frozen.
-
-
Are the two servers plugged into the same power source or different ones? Do they freeze at the same time, or does it happen at different times?
-
After a crash, try to SSH into the server and see if you can recover any logs by running this command:
sudo journalctl -b -1
Let us know, and we’ll take the next steps from there.
When the server freezes, it will not ping. SSH does not function because the comms are down. The servers don’t shut down at the same time. Common power supply not a problem. Cycling power has so far always been successful. BUT the logs are rebuilt and the pre crash data is deleted
It’s odd that you have two servers and they both have an issue that we’ve never seen before, this is why we’re asking questions to see if there’s something specific to the environment you’re in that is a contributing factor.
Like for example, they’re both plugged into to a surge protector or UPS – in which case we’d ask you to connect one to another power outlet elsewhere in the home and monitor. There are only so many questions like this that we can ask, and things about your environment we can guess at, so we’re reliant on you (knowing that we’ve never seen a Server Pure lose comms and not also power off) to help us determine anything we might not think of.
Here’s yet more things to check…
-
When you say, lose comms, do you also mean it’s not connected on the router and it’s also not ping-able? You didn’t answer @h0mer 's questions there.
-
When the servers can’t be reached, are the ethernet port lights on?
-
You say Kiosk mode is the same, have you been able to catch what it looks like? Is the screen just black?
-
Take a look at the power supplies, look at the output. Do you have another at home that’s the same output? They’re pretty standard, almost all mini PCs and laptops have the same output. It wouldn’t be impossible that the power supplies are bad, we’ve seen a few of those lately.
I’m sorry for not being clearer, I’m an electrical guy so comms means the network is down. Ping to the ip fails, the lights on a switch nearby stop blinking that are for the server. And this is like the power supply problem. Both servers are connected to one ups, but since the problem has never happened to both servers at the same time I have not split the power. Actually I originally only had trouble with one server but now with the new power supplies I have had both freeze up. But not as often. I’ll look around for some bigger charges.
But, ( I keep asking) can you tell me a way to permanently and continuously save the log files even if the power is the cause
When these servers freeze up, One time I connected a monitor mouse and keyboard. The monitor does not get a signal. I’m lazy now, when I cannot log on with the same laptop that was working before, I ping the ip, run angry scanner, and if it’s gone. I cycle power, and everything works after it is on.
Understood. We’ll get this sorted, I just need to go through the evidence collection process to get some idea as to where to begin.
One further thing I’d love for you to try before we bring these in to look at is for you to experiment with the power supplies. You said you already replaced them and that didn’t help? We have seen a few issues with power supplies recently and it has been the first place I’ve looked in the last few months. Since you have the experience necessary for this, your opinion would be valuable. Maybe you have some power bricks around?
While this all sounds like a hardware problem, and perhaps a power supply problem (it’s just so unlikely that two machines have the same issue, while a batch of power supplies being bad isn’t impossible) if it’s a software issue, we can find it with this…
sudo journalctl -k -b-1
and
sudo journalctl -b-1
But it your case this hasn’t shown anytihng