Portal Home > Knowledgebase > Articles Database > Identifying VPS problems
Identifying VPS problems
| Posted by ronaldst, 02-01-2016, 01:42 PM |
|
I am paying for a managed VPS with WHM/cPanel. Until recently I've had no issues but some minor hickups with some updates (had to be done manually).
Recently I've been struggeling with complete lockout/hogging though, and this is without any "worthy" changes to the VPS or the websites hosted. As I got very limited experience I am unable to identify the cause of the problems myself.
What happened..
Yesterday evening the VPS stops responding. It hogs and I am unable to grab mail, browse any webpages or domains on the server, cannot connect to SSH and even the hosting company "VPS Control Panel" use forever to show the status of the VPS to offline. It's clearly affected by a VPS not replying to the hosting company's web interface. For a while it did respond to PING (1-2 hours on both IP's). After that it went completely off the radar.
The hosting company spent like a day to fix it, even though they kept telling me they were working on it "full force". So it came back up, stays up (idle) for about 30 minutes until the same thing happens again. It's not responding to anything but PING. Just seem very odd to me.
I asked what was the problem several times, but either they don't want to explain it to me (as in I dont need to know) or they simply won't tell me the problem for some other odd reason.
It's been hours since support were again "back on it full force" without any progress.
What can it be? The VPS is running "LAMP", 2 websites/domains with WHM/cPanel.. Solid hardware with 3 Core, 2GB RAM, XEN based.
How do I identify what might be causing this? I've lost a little faith in the "management" at the moment and thinking of taking things into my own control.
Last edited by ronaldst; 02-01-2016 at 01:52 PM.
|
| Posted by nessa, 02-01-2016, 03:39 PM |
|
I'm not sure that many of us here will be able to tell you what happened without having looked at your server. If your VPS is managed, your best option would be to contact your host. It sounds to me like they probably had an outage, which does happen. I can tell you that in general, hosts try to be vague about the technical details of specific problems that may be widespread and affecting multiple people. Most will take the posture of fixing the issue first and worrying about PR later. So it's possible they either don't quite know what the problem is, haven't yet decided what to disclose, or just don't want you to know what happened. But honestly, I'm not sure how any of us would be able to help you with this - anything we suggest would be nothing more than a guess.
|
| Posted by MightWeb-Greg, 02-01-2016, 04:25 PM |
|
To know of Echo with nessa was saying... We're not going to be able to tell you what is going on with your VPS without taking a look at it. They may or may not be able to fix the issue.
If you think you can help fix it or resolve the issue I say why don't you give it a try? The worse thing to come of it is you'll have a fresh backup and you may even be able to resolve the issue. Unless it's something like a node problem or an attack they're facing.
|
| Posted by ronaldst, 02-01-2016, 05:51 PM |
|
I don't know I might just be overthinking this. Got a really weird feeling that something is not as glamourous as presented in the marketing.
I've been told the node hosting my VPS is overloaded and they are monitoring the situation. Thats the status as of right now, the VPS and its hosted websites are timing out still.
The CPU/MEM is dedicated to my VPS right? So any load info I'm seeing in 'uptime' or 'top' is purely mine? It's way too high right now. Could it be lack of cpu power, not cores? Poor technology or infrastructure? I don't understand the technical aspect here but 3 Cores/2GB MEM seem sufficient for a Wordpress site to me..
|
| Posted by nessa, 02-01-2016, 06:09 PM |
|
Again, these are questions you should be asking your host. I can tell you that in a typical VPS environment, at least the memory would be dedicated and often a certain number of cores as well. In some environments though (like VZ), the load you see when running 'top' could be affected by activity on the parent server. In any case, if a physical server is overloaded it could cause stability problems with the VMs it hosts. I could take guesses at what it is, but that won't help you. If it's that big of an issue, call your host up and ask them to move your VPS to another node.
|
| Posted by ronaldst, 02-01-2016, 06:29 PM |
|
VPS is Xen -based. Again, not sure of the technical details of that, but they have come out and told me the node is overloaded and they are working on a solution for me. I assume that is either moving me or someone else to a different node. I have a problem with patience though, and this is taking seriously long time.
|
| Posted by HostXNow_Chris, 02-01-2016, 06:34 PM |
|
Who is the provider? What did they say about this?
|
| Posted by net, 02-02-2016, 12:22 AM |
|
How long you've been waiting now? If this is a cheap vps, then expect an overloaded node. Can't say more....
|
| Posted by HWH-Jessica, 02-02-2016, 12:39 AM |
|
Definitely looks like a problem with the main node. Is the support team helpful? Most importantly, do you have your own backups? I guess the better option would be to ask the host to move your VPS to another node, which should solve the issues for the time being.
|
| Posted by ronaldst, 02-02-2016, 07:16 AM |
|
To be fair I'm not getting much answers from my host. It's been 36+ hours now, and since last night I haven't heard a thing from support. The VPS have been on/off all through the night and it seem to be going under everytime it has visitors. 'uptime' and 'top' show loads hitting 200+, that cannot be normal on a XEN 3CORE/2GB MEM VPS. LFD is pushing out emails of services running failing, excessive resources and out of memory.
This is why I did get managed hosting. To sort the server and point me in the right direction if anything is off. Isnt that the basics in "management"? I don't want to name and shame the host. It's a $40/month XEN VPS 3CORE/2GB MEM/80GB (non SSD). Not sure if that is considered cheap or not, but they did present themselves well and it have felt solid up until now.
I have complete backups and am thinking about it, hard.
Anyway, thanks for the input guys.
Last edited by ronaldst; 02-02-2016 at 07:20 AM.
|
| Posted by ronaldst, 02-02-2016, 08:43 AM |
|
So I've just gotten a reply from their support team about my ongoing ticket. The VPS is as explained in the post still loaded with periodical problems. I requested a new status some information about the specifics of the problems going on to my VPS, and I also asked if moving me to another node was a solution.
As I am reading this the VPS is unreachable by any form and means other than PING. How do you respond to these things? As far as they are concerned it seem to be acceptable to wait for these problems to resolve themselves.
I dont know what to do other than look for a new host, really. Maybe (if I am lucky) get a refund of remaining months of service.
Last edited by ronaldst; 02-02-2016 at 08:47 AM.
|
| Posted by HostingBig, 02-02-2016, 10:05 AM |
|
well if you are absolutely 100% sure the issue is not on your server (which if the host has all ready took responsibility) then you have no choice but to move. Sounds to me like they have no clue at this point or they have no where to move you.
|
| Posted by ronaldst, 02-02-2016, 11:42 AM |
|
I got some good news in all of this. It does raise quite a few question marks though but the VPS is back up and responds well. Loads are very high still though and services like cpsrvd are failing left and right, PHP causing excessive resource usage, memory outage and so forth. I think that is weird though, considering it has been running fine prior without these issues.
I have never really disqualified the possibility the issue is on my server. That is kind of why I made this post, I need to get a better understanding of why loads are high. Simply identifying problems that occour.
However, I highly doubt my single Wordpress site should cause these kind of issues with a VPS of these specs. I would very much expect a 3CORE/2GB MEM VPS to run a small Wordpress site flawless. The coding is very well done and analyzing the plugins and load times show nothing out of the ordinary. Even my old shared host with CPU, MEM and IO -limits had no problems..
Last edited by ronaldst; 02-02-2016 at 11:55 AM.
|
| Posted by BoxIntense, 02-02-2016, 09:44 PM |
|
Glad to hear that it's been fixed.
Considering the filesystem has been corrupted, I would assume that there are other parts of the systems that might get corrupted as well, for example cPanel's core binaries. Therefore it would be best to run upcp --force as well. If everything doesn't work, then your best bet would be a complete re-install of the OS. Good luck
|
| Posted by ronaldst, 02-02-2016, 09:57 PM |
|
I don't know where to begin, so many things have happened since my last post.
To be short - The good news did not last long. The VPS was working somewhat an hour or so until it again hogged and went down. The funny thing is that once I report it to their support, they wait answer until the load is lower and the VPS is responsive again.. even if it takes several hours. Same happened this time around.. I'm wondering if they got PING running and once they see replies they answer the ticket: "Everything is back up and running. We did measures and patches, should be fine now but we are monitoring". That seem to be the casual answer I'm getting.
If it was the file system or not I don't know. I doubt it but I'm no technician. The VPS went back down about an hour later, then back up for 10-15 minutes (after 1-2 hours downtime). Thats how it has been since the problems first started. Up, down, up, down. Mostly down though.
I asked them to give me some guidance on how to cancel the service and possibly a refund of some sort. No reply to that question, they avoid it as any other specific questions I've asked. They keep telling me the server is ok.. even though it's mostly down Last couple of hours I've been getting about 800 emails from the LFD reporting anything from excessive loads to corrupt databases.
I'm thinking of copy/pasting the whole support ticket conversation to this post. Name and shame. There is definitely some entertainment value. These guys are obviously lost in managing these servers.
I'm currently trying out 2 new VPS companies, I'm moving either way. Tired of this ********.
Last edited by ronaldst; 02-02-2016 at 10:01 PM.
|
Add to Favourites 
Print this Article
Also Read
Shared Hosting
Shared Hosting
