microLXC Public Test

TheOnlyDK

@jcn50 said:

@TheOnlyDK said: Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

I have absolutely no issue with the Alpine image @ Melbourne!

If only I knew how to use Alpine lol

TheOnlyDK

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

I reinstalled the server and just let it ran. Got an email telling me I've used too much CPU in the last 30 mins and it's still stuck there. I'll cancel the operation now to prevent any other complications.

Neoon

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

TheOnlyDK

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

Port forwarding is fine now. I'll try spinning up another server in Melbourne, see if that'll do anything. When it "hangs", I can't actually kill it with CTRL-C. Not exactly sure what it's actually doing. If I just let it run for however long it wants to run, would that be a problem? I might get a few high cpu alerts, but any "punishments" for that?

Neoon

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

Port forwarding is fine now. I'll try spinning up another server in Melbourne, see if that'll do anything. When it "hangs", I can't actually kill it with CTRL-C. Not exactly sure what it's actually doing. If I just let it run for however long it wants to run, would that be a problem? I might get a few high cpu alerts, but any "punishments" for that?

You can list processes with ps ax and kill them with pkill.
The "punishment" is a restart after x hours, because most of our resources, are also shared not dedicated.

TheOnlyDK

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

Port forwarding is fine now. I'll try spinning up another server in Melbourne, see if that'll do anything. When it "hangs", I can't actually kill it with CTRL-C. Not exactly sure what it's actually doing. If I just let it run for however long it wants to run, would that be a problem? I might get a few high cpu alerts, but any "punishments" for that?

You can list processes with ps ax and kill them with pkill.
The "punishment" is a restart after x hours, because most of our resources, are also shared not dedicated.

Debian 12 appears to be working fine, no issues with apt-get. I wonder if the kernel is causing some issues with Debian 13? I don't know...

Neoon

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

Port forwarding is fine now. I'll try spinning up another server in Melbourne, see if that'll do anything. When it "hangs", I can't actually kill it with CTRL-C. Not exactly sure what it's actually doing. If I just let it run for however long it wants to run, would that be a problem? I might get a few high cpu alerts, but any "punishments" for that?

You can list processes with ps ax and kill them with pkill.
The "punishment" is a restart after x hours, because most of our resources, are also shared not dedicated.

Debian 12 appears to be working fine, no issues with apt-get. I wonder if the kernel is causing some issues with Debian 13? I don't know...

Containers are sometimes a bit fragile, I am actually thinking about switching to KVM.
Which obviously doesn't come without sacrifices, some locations will get deconsecrated.

TheOnlyDK

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@TheOnlyDK said:

@Neoon said:

@Neoon said:
Expect a few reboots, I found a potential fix for this.

Should be fixed on all nodes, except 2.
Lemme know if you have any issues.

Melbourne appears to have some bug where apt stucks at "Processing triggers for shared-mime-info". This is on Debian 13 image, no issues at other locations. Server doesn't appear to be hung, can still type into the console.

Checked that, had no problems on a test deploy.
If apt-get is stuck then you can kill it and run it again.

Auckland appears to have broken port forwarding on ipv4. I don't have access to ipv6 at the moment to test that, but SSH doesn't work on default port or any other ports assigned. This is on Debian 13 image. SSH works fine at other locations once the package is installed.

Should be fixed.

Port forwarding is fine now. I'll try spinning up another server in Melbourne, see if that'll do anything. When it "hangs", I can't actually kill it with CTRL-C. Not exactly sure what it's actually doing. If I just let it run for however long it wants to run, would that be a problem? I might get a few high cpu alerts, but any "punishments" for that?

You can list processes with ps ax and kill them with pkill.
The "punishment" is a restart after x hours, because most of our resources, are also shared not dedicated.

Debian 12 appears to be working fine, no issues with apt-get. I wonder if the kernel is causing some issues with Debian 13? I don't know...

Containers are sometimes a bit fragile, I am actually thinking about switching to KVM.
Which obviously doesn't come without sacrifices, some locations will get deconsecrated.

KVM is nice but I kinda enjoy containers. It's fast and efficient, none of the bloat overhead from KVM.

Carlin0

Vilnius and Bratislava have been experiencing issues with IPV6 for a couple of days

Neoon

@Carlin0 said:
Vilnius and Bratislava have been experiencing issues with IPV6 for a couple of days

The IPv6 checks used to run via UptimeRobot, however, it looks like some of them don't have a email attached at all. I added them 100%.
Because of legacy reasons IPv6 wasn't added to Hetrixtools.

Usually I use HetrixTools + UptimeRobot but on IPv6 its only UptimeRobot.
Plus the new interface is pure cancer, I have to find another solution for this.

I check on it when I got time.

Neoon

@Carlin0 said:
Vilnius and Bratislava have been experiencing issues with IPV6 for a couple of days

Should be resolved.

Carlin0

@Neoon said:

@Carlin0 said:
Vilnius and Bratislava have been experiencing issues with IPV6 for a couple of days

Should be resolved.

Yes, thank you, they're back

kadidala

f4c8-452f-376e-d6ca
thanks so much

Neoon

@Neoon said:
Maintenance announcement
Sandefjord and Helsinki gonna get migrated to Incus next Friday.
Expect up to 60 minutes of downtime, should be way shorter but will see.

I am targeting 21:00 CET, subject to change though.

Done, Downtime for the migration itself was only a few minutes.
Except for Virtual Machines in Sandefjord, there was a bug that had to be fixed first.

All should be online now.

kadidala

@kadidala said:
f4c8-452f-376e-d6ca
thanks so much

hello，
My dashboard is currently showing "Quota exceeded" and "Quota at zero (0MB)", and I haven't been able to access for container.
Could you please check this or raise my quota.
thanks

d2411

e406-6810-edeb-5f71

d2411

c292-1ee9-53f8-28a5

Sharmaishaan72

05bc-6951-040e-e950

Neoon

Helsinki is running a bit hot currently due to a failed CPU fan.
I have no ETA on a replacement yet, new Deployments are disabled for now.

Neoon

@Neoon said:
Helsinki is running a bit hot currently due to a failed CPU fan.
I have no ETA on a replacement yet, new Deployments are disabled for now.

The CPU fan isn't likely going to get replaced, please backup your data.
We should get a replacement dedi soonTM.

World

Islamabad has down for a week, any update?

Neoon

@Neoon said:

@Neoon said:
Helsinki is running a bit hot currently due to a failed CPU fan.
I have no ETA on a replacement yet, new Deployments are disabled for now.

The CPU fan isn't likely going to get replaced, please backup your data.
We should get a replacement dedi soonTM.

CPU fan got replaced, server is back online, deployment should be available shortly again.

Neoon

@World said:
Islamabad has down for a week, any update?

There was an IP swap, as announced a few weeks ago.
HetrixTools wasn't updated yet.

World

@Neoon said:

@World said:
Islamabad has down for a week, any update?

There was an IP swap, as announced a few weeks ago.
HetrixTools wasn't updated yet.

I know there was an IP swap and I've changed when I saw the announcement some days ago, but currently the new IP is unreachable and from the microLXC panel it returns

Failed to contact Node, please try again later.

Neoon

@World said:

@Neoon said:

@World said:
Islamabad has down for a week, any update?

There was an IP swap, as announced a few weeks ago.
HetrixTools wasn't updated yet.

I know there was an IP swap and I've changed when I saw the announcement some days ago, but currently the new IP is unreachable and from the microLXC panel it returns

Failed to contact Node, please try again later.

I ticket them, something broke when they removed the old IP.
IPv6 is also entirely gone.

Neoon

Yesterday I forgot to disable the offloading since the fan replacement in Helsinki.
Today the NIC died, due to this, the script however, recovered the machine within 15 minutes.

Fixed.

Neoon

@Neoon said:

@World said:

@Neoon said:

@World said:
Islamabad has down for a week, any update?

There was an IP swap, as announced a few weeks ago.
HetrixTools wasn't updated yet.

I know there was an IP swap and I've changed when I saw the announcement some days ago, but currently the new IP is unreachable and from the microLXC panel it returns

Failed to contact Node, please try again later.

I ticket them, something broke when they removed the old IP.
IPv6 is also entirely gone.

Got a ticket reply, they escalated it, no ETA though.

Neoon

@Neoon said:

@Neoon said:

@World said:

@Neoon said:

@World said:
Islamabad has down for a week, any update?

There was an IP swap, as announced a few weeks ago.
HetrixTools wasn't updated yet.

I know there was an IP swap and I've changed when I saw the announcement some days ago, but currently the new IP is unreachable and from the microLXC panel it returns

Failed to contact Node, please try again later.

I ticket them, something broke when they removed the old IP.
IPv6 is also entirely gone.

Got a ticket reply, they escalated it, no ETA though.

Still no solution yet.

david

I think the bandwidth meter for the control panel is stuck, at least for Helsinki. I checked it on January 1, and it was 74.7 MB, and again today, and it's still 74.7 MB. I have vnstat running also, and it is increasing.

Neoon

@david said:
I think the bandwidth meter for the control panel is stuck, at least for Helsinki. I checked it on January 1, and it was 74.7 MB, and again today, and it's still 74.7 MB. I have vnstat running also, and it is increasing.

The cronjob was stuck, Validvia had a dsync, I cleaned up but I forgot one thing, hence it was stuck, fixed now.