After the VServer had been running without problems for a few months, there have been problems since last night.
Around midnight, the first message came from the monitoring system that the server was not accessible. After that, messages came every half hour.
So this morning I went to the notebook to see what was going on. It quickly became clear that the server was not accessible, ping did not work, websites could not be called up. Well, maybe it was another problem that something had opened up. But that rarely happens with Linux itself.
Like last time, I wanted to switch to the rescue system and then restart the server normally. Unfortunately, it has been stuck in the rescue system ever since.
Already during the last failure, the switch between rescue systems took quite a long time, as it is also described here. Normally, the process should take 10 20 minutes. I waited a good 2.5 hours and then called Strato.
After a good 10 minutes on hold, someone was on the spot. The employee was able to narrow down the cause and mentioned problems with the “hardware node” for the V-Server. A ticket has been created and now we have to wait and see.
Second call
Still no improvement in sight in the evening, time for a second call. The employee is nice, but can’t really help. There is already a ticket and it should be dealt with at some point. Unfortunately, he can’t say when that will happen. In addition, he recommends that the VServers should not be used for productive systems anyway. The question is, what else are they for?
Strato vs. Hosteurope Support
In direct comparison to our web hosting with Hosteurope, Strato support is rather poor. With Hosteurope, if there was a problem, it was enough to send an email via the HIS system and within an hour the problem was usually solved and the server restarted.
Update after 24 hours
No improvement even after 24 hours. The server continues to hang in recovery mode. Since this is not a state of affairs, I already restored the backup from the server at Hosteurope yesterday, bent the DNS around and our websites are accessible again.
Update 2: Server up and running again
A good 30 hours after the outage, the server was running again, there was no feedback from Strato. I happened to see in the admin area that the rescue system had now been restarted. I then restarted the normal system, which worked without any problems within a few minutes.
Cancelled
The end of the story was that I backed up and downloaded the websites and databases again. I then re-installed the data onto the Hosteurope webspace. The original backup was a few days old, so I was up to date again.
I then cancelled the VServer. The positive thing is that this can be done easily via Strato’s web interface. The notice period is only one month.
Conclusion
In principle, it is not a problem if problems occur. However, they should be solved as soon as possible. If a server is not accessible, it should be restarted somehow within a short time. Strato’s support was very helpful, but they couldn’t open more than one ticket. Whether the ticket was processed or not, could not be found out until now. The server was running again, but there was no response to the ticket.
