VPS up, down and exposed?

[anthonyD]

Hi

I’ve been with EUK a couple of months

Problem 1
My VPS was on the “Windows Node Problem” node with a failed array so disappeared for a couple of days. History in the network status thread.

Problem 2
On Saturday all the sites on my VPS stopped including plesk, RDP and mail continued to work fine, this lost us a full days sales not to mention the ad-word costs driving customers to a dead site (yes I could have disabled them but I didn’t expect a 7 hour outage).

The windows admin finally came back and told me this was an IIS issue on the main node which was now fixed.

Problem 3
On Monday all the sites stopped working, again RDP and mail continued fine. Initially the admin blamed virtual memory which he said had been increased. Looking at the memory in task manager this appears unchanged so not sure what this was about.

Problem 4
The sites worked briefly after the reboot then again all the sites stopped again. After investigation I was told the problem was caused by the firewall being switched on and the admin has disabled it!

In the VPS knowledge base is clearly states that windows firewall is a MUST, my VPS currently has this switched off – which is the correct setting for VPS?
Is the VPS protected by a network firewall or is my VPS just sitting there open?

I bit of clarity would be appreciated.

Thanks

[eUK-Nick]

Hello Anthony,

Normally windows firewall is turned on. But at times, it hampers connectivity to the VPS from client side because of which it is turned off for that particular VPS. I have asked our VPS admin to look into your windows VPS right away. I will update you in an email once I have more updates.

[eUKShane]

Hello,

Sorry for the delay in response.
The node on which your VPS is hosted is up and running from last 14 days. It was rebooted 14 days before because of instant Power failure problem, but it has been already sorted now.

>> The windows admin finally came back and told me this was an IIS issue on the main node which was now fixed.

There may be some different reason, because VPS don't refers hardware nodes services and IIS is completely disabled on the hardware node.

Regarding Windows firewall, you have to configure it properly and then you can use it. We are already having Network Firewall and Antivirus installed on main Hardware Node.

If you wish, we can migrate your VPS to our new Hardware node having same configuration. We will migrate your VPS as it is, but you will get different IP address. You can place a ticket or email to vps@eukhost.com if you would like to proceed with it.

[anthonyD]

Hi

Your answer has me concerned for a number of reasons.

If you look in the network status forum at the Windows Node Problem you will see that the node suffered an array problem with the first entry on 17th at 9:51am with the fix being notified on the 18th at 9:17 both by Nick.

If IIS is disabled in the hardware service and VPS doesn’t refer to it anyway why did the admin site this as the problem?


Could you expand on how to configure the firewall properly please as http port 80 open is enough to allow IIS to deliver content?


>> We are already having Network Firewall and Antivirus installed on main Hardware Node.

I don’t understand what you are saying.
Does this mean you have an independent network hardware firewall (eg Cisco,etc..) and have AV installed on the main node on which the VPS containers are hosted? If this is the case windows OS level software firewall in not needed.


Please explain what benefit will I gain in moving the VPS to a new node?
Given that new IP address would require a DNS update which could put my sites out of action while the DNS propagates up to 48hours.

Sorry I have replied with more questions but my concern is that an underlying problem has not been fixed and my sites could go off-line again.

Thanks

[eUK-Nick]

Hello Anthony,

Let me elaborate on this to you.

1. The RAID Array problem which you are referring to has been completely fixed. Virtuozzo uses only a few selected components of IIS and not the whole web-server itself. On the contrary, having a full fledged IIS service on the node is not a very good security measure.

2. All our nodes have Dr.Web AV installed on them. We also have Windows firewall on the nodes and a couple of other security measures which protects it against Denial Of Service attacks and similar attempts to disrupt its performance.

However, this does not actually mean that you can have Windows firewall turned off for your VPS. The OS on your virtual machine is independent of the main node and also has its own private area. When a new VM is created, the windows firewall is turned off by default. It is upto the user to turn it on or have some other security measure as per their usage and needs.

3. As of now, your VPS does not require a move to a different node. The current hardware node is functioning without any issues.

If you would like to have the windows firewall configured for your individual VPS (which we do recommend), you may contact support over live chat or email the VPS team. They will straight-away do it for you.

If you have any more concerns, I would be glad to address them.

[jc8654]

Just to point out as well for your information about migration, there wouldn't be any downtime as you would set the current DNS up so it forwarded to the new server and therefore during the propogation time your site would still be accessible. Obviously you don't need this now but thought it may be of use should you need it in the future.

[anthonyD]

Hi

All our sites inc plesk are not responding again, RDP and mail continue to work OK.

This is really becoming untenable, I’m rapidly loosing any confidence in the quality of VPS hosting at EUK.

I’m on with live chat as well to try get this sorted.

And being told that the hardware node is being rebooted now. So all VPS will be down…!



Given the comment above about DNS forwarding, this suggest that a clean VPS install may be worth trying.

I assume that EUK would need to manage this on the nameservers.

Please HELP!

[eUK-Nick]

The node is up. All the VPS' are running including yours. The virtuozzo service got stuck which had to be refreshed. This has been done. Everything is normal now.

[anthonyD]

The live chat team seem happy to just reboot either the VPS or the hardware node. So all customers on the same node will have had an outage this morning!

Just rebooting isn't getting to the root of the problem though. Or is it just that the SWSoft VM/VPS are a bit britle?


Short of a full dedicated server can I get a resonably reliable service?


Anthony.

[jc8654]

I have two VPS with eUK and although I do occasionally have problems, they are no more than I would with a dedicated server. Virtuozzo sometimes seems to have a mind of its own when doing things so I'd wait and see how things progress now that it's hopefully sorted and if it is still a problem then maybe think about migrating to a different node.

[eUK-Nick]

Originally Posted by anthonyD

The live chat team seem happy to just reboot either the VPS or the hardware node. So all customers on the same node will have had an outage this morning!

Just rebooting isn't getting to the root of the problem though. Or is it just that the SWSoft VM/VPS are a bit britle?


Short of a full dedicated server can I get a resonably reliable service?


Anthony.

We have several windows semi-dedicated and VPS nodes which run very well. Yes, virtuozzo does have its own problems. At the end of the day, it is a semi-dedicated environment which is working on the concept of virtualization. This is not to say that its not reliable and cost effective. On the contrary, they are very much so, but a little more patience is required when working with these.

With regards to the specific node on which your virtual machine is setup, you would not have any more problems. The problems with Virutozzo have been cleared. Thank you for your understanding.