OVH Community, your new community space.

(The mks thread) Anybody else get DEFECT and END OF DEFECT emails


heise
01-03-2015, 20:31
Like an hour.

netbeamerp
01-03-2015, 15:43
Hi,
My Server: RBX2 - Rack: 23B06 - Server ID: 162393

i am getting this message on
{ "message": "An intervention is being carried out on this server, at the moment you can't perform a hard reboot yourself" }

how much time it will take to get server back.
very poor support.

mks
16-02-2011, 10:42
Seems like I will be assigned a new advisor.

Hoping new advisor will resolve this for me.

mks
16-02-2011, 10:33
Can somebody here ask Dan at the IE division to look at my ticket?
Ticket #656809

I've been staying up all night waiting to get a response and the responses I got from the advisor were ridiculous.

I cannot connect to SSH and cannot reboot the server via the OVH manager yet the advisor tells me to install SSH key? How am I supposed to do that when I cannot connect to SSH?

After telling the advisor that I cannot REBOOT and cannot get into SSH so I CANNOT INSTALL THE SSH KEY, the advisor tells me to go into RESCUE MODE?

If I cannot REBOOT the server, HOW CAN I GET INTO RESCUE MODE?

Does this type of support seem acceptable?

mks
09-01-2011, 06:24
Can OVH staff look at my server at all?

I paid for bandwidth 2TB but speed is still limited to 10mbps even though it says I have 1100GB remaining! The staff's suggestion of reboot via manager did not work either!

mks
27-12-2010, 18:46
Does OVH open today or still holiday?

Waiting for my order since 3 days ago:
Bandwidth order 8980463

And other orders:
Bandwidth order 8991874
New server order 8991877

Neil
08-11-2010, 10:33
Hi

As it is US Order it will be processed when IE Staff are available, as this has repeated happened to you, maybe you should not pay at weekends but instead on weekends.

Your order has been processed now, but I have disabled you from making new threads. You can now only reply to current threads because of keep opening pointless threads,

jonlewi5
08-11-2010, 10:18
Then email support!

These forums are not official support methods. It may be worth contact the .ie support aswell.....

mks
08-11-2010, 09:59
Finally they process my bandwidth order but what is this???
Consumed this month : 7.25 TiB
Remaining this month : 1.07 TiB

I should have 1.75 TiB remaining.

mks
08-11-2010, 03:20
Quote Originally Posted by Myatu
A British one, perhaps. But not an Irish one.

Don't keep expecting the UK staff to go out of their way to help you each and every time you decide to order things just before the weekend starts, especially since they're not responsible for US/IE customers...

It irks me, because UK support isn't particularly well staffed that they can cater each and every mundane request from every single country in the world. I'd like them to spend their time and effort on UK customers instead; that's what they're here for.
UK staff can inform IE staff that there is a pending order. It doesn't take that long for them to do this and I am sure communication from OVH staff to OVH staff have higher priority.

It has now been over 50 HOURS since I paid for the extra bandwidth and it still has not been processed. By now, I probably have used at least 600GB of bandwidth at 10mbps. Why isn't this an automated process?

yonatan
07-11-2010, 02:30
Quote Originally Posted by Myatu
A British one, perhaps. But not an Irish one.

Don't keep expecting the UK staff to go out of their way to help you each and every time you decide to order things just before the weekend starts, especially since they're not responsible for US/IE customers...

It irks me, because UK support isn't particularly well staffed that they can cater each and every mundane request from every single country in the world. I'd like them to spend their time and effort on UK customers instead; that's what they're here for.
+1 on that.
Exactly what i had in mind.

Also for a bandwidth order , you can submit an incident ticket about it.
Select your server, then
under "My problem is related to:"
select "Order Being Processed / Error "

they do help!

Myatu
07-11-2010, 00:58
Quote Originally Posted by mks
What was I thinking posting at an OVH forum hoping an OVH staff would read my post?
A British one, perhaps. But not an Irish one.

Don't keep expecting the UK staff to go out of their way to help you each and every time you decide to order things just before the weekend starts, especially since they're not responsible for US/IE customers...

It irks me, because UK support isn't particularly well staffed that they can cater each and every mundane request from every single country in the world. I'd like them to spend their time and effort on UK customers instead; that's what they're here for.

Razakel
07-11-2010, 00:51
Quote Originally Posted by mks
What was I thinking posting at an OVH forum hoping an OVH staff would read my post?
You misunderstand. OVH.ie staff don't read or post on the OVH.co.uk forum, so you're posting in the wrong place.

mks
07-11-2010, 00:24
What was I thinking posting at an OVH forum hoping an OVH staff would read my post?

Myatu
06-11-2010, 23:37
Quote Originally Posted by mks
OVH staff can read these forum posts and process the order, not the wrong forum.
I'd like a Texican Whopper meal, medium and a coke. Eat in, please. (hoping a Burger King employee happens to read this...).

mks
06-11-2010, 20:23
This is the most active forum for OVH, I am from United States customer of IE division.

OVH staff can read these forum posts and process the order, not the wrong forum.

yonatan
06-11-2010, 19:22
Euros?
are you posting in the wrong forum?

mks
06-11-2010, 19:22
By Monday, the 2TB extra bandwidth I bought for 30 Euros would have been nearly gone while the server is running at 10mbit!

MicroChip123
06-11-2010, 18:27
OVH UK staff have to do it manually and they dont work weekends so expect it monday.

mks
06-11-2010, 18:20
Quote Originally Posted by DigitalDaz
As john shows above, it is very easy to turn off.

Though it may not be so effective in that it sometimes reports a server as down when it is not, is that it will never report as server as UP when its not.

For most of us I would imagine this is a good thing as it then schedules an intervention,
I just turned off monitoring for all servers, instead I have it monitor port 80, it's much more effective imo. If the server is truly down, I will do a hard reboot, if it won't come back up, an intervention will be scheduled automatically.

Because even if tech performs an intervention when I am not monitoring the server, once the server respond to pings, the services will be all down anyway for several hours. It is useless to me.

Quote Originally Posted by Razakel
A server not pinging indicates something bad has happened, as opposed to services not responding, which could be anything.

It automatically schedules an intervention, so a tech will go and have a look at the server and then probably reboot it, or schedule a hardware replacement if necessary.
In my case, the server is always not available for several hours after hard reboot.

mks
06-11-2010, 18:13
I paid over 17 hours ago but still not processed yet.
I need server back to 100mbit.

Order #8665533

Razakel
06-11-2010, 12:06
A server not pinging indicates something bad has happened, as opposed to services not responding, which could be anything.

It automatically schedules an intervention, so a tech will go and have a look at the server and then probably reboot it, or schedule a hardware replacement if necessary.

Winit
06-11-2010, 11:03
Quote Originally Posted by LawsHosting
What happens if you disable pings (icmp) to the server when monitoring is on? It must confuse the monitoring systems, obviously.
It does.

LawsHosting
06-11-2010, 10:59
What happens if you disable pings (icmp) to the server when monitoring is on? It must confuse the monitoring systems, obviously.

DigitalDaz
06-11-2010, 10:07
As john shows above, it is very easy to turn off.

Though it may not be so effective in that it sometimes reports a server as down when it is not, is that it will never report as server as UP when its not.

For most of us I would imagine this is a good thing as it then schedules an intervention,

jonlewi5
06-11-2010, 09:12
manager > server status > monitoring > turn off.

mks
06-11-2010, 08:26
Sometimes when my server has high load, I will receive DEFECT email that server is not pingable and an intervention will take place.

Please tell me, when a server is not pingable, what does the OVH monitoring program do? Does it automatically do a hard reboot? What else does it do?

Because I find that whenever I get that email, I have to go into rescue mode and check the hard drives and then wait a few hours before the server comes back online.

If the server is being sent a hard reboot request after not responding to ping, how do I opt out of this monitoring? I would rather wait until the server load to come down in 10 minutes instead of waiting HOURS for the server to come back up after the hard reboot if that is what's happening.

I have servers at many other providers and OVH is the only provider I keep having this problem. The other providers don't monitor pings, etc. though since I usually find out the server is down on my own. But I would rather do that than wait upwards of 11 hours to get my server back up after a high load.

Please respond of how to opt out of ping monitoring.

The ping monitoring is not very effective anyway since the server responds to pings but SSH and all other services are down. This gives false belief that the server is running fine when in fact it is not!

mks
03-11-2010, 06:52
USA server to OVH server (1.5 Mb/s) Still pretty poor speed
OVH server to USA server (150 Kb/s) Horrible

On another USA server I would get better speed but still poor:
USA server #2 to OVH server #2 (5.6 Mb/s)
OVH server #2 to USA server #2 (439 Kb/s)

Andy
02-11-2010, 12:15
Probably not great because it uses transit links which tend to be overloaded. Soon OVH will have their peering fully set up so there may be some increase then.

jonlewi5
02-11-2010, 10:53
yep you can just let it expire, or you have the option in the manager to return the server.

mks
02-11-2010, 10:49
And is there a huge difference between 100mbit and 1gbit connection to and from USA? I tested from the bandwidth speed test thread, I get the same download speed (very slow 100kb/s) from a 100mbit 1gbit and 10gbit server?

mks
02-11-2010, 10:46
If I need to cancel a server, do we NOT renew it or we need to email support?

mks
01-11-2010, 16:54
I got it now.

Andy
01-11-2010, 12:45
Quote Originally Posted by mks
Sorry I still don't understand.

If I upload my 5GB site to my OVH server from home (Internet to OVH), is that 5GB counted as bandwidth?
No. Only outgoing bandwidth from your server (your server uploading) is counted. All incoming bandwidth is free. All OVH to OVH bandwidth is free regardless of upload or download. Only outgoing bandwidth outside of OVH's network is counted.

Neil
01-11-2010, 12:43
Quote Originally Posted by mks
Sorry I still don't understand.

If I upload my 5GB site to my OVH server from home (Internet to OVH), is that 5GB counted as bandwidth?
No, only data that leaves the server and goes outside the datacentre(s) is counted.

mks
01-11-2010, 12:18
Sorry I still don't understand.

If I upload my 5GB site to my OVH server from home (Internet to OVH), is that 5GB counted as bandwidth?

Andy
01-11-2010, 12:12
External bandwidth is any data NOT coming from inside the OVH datacenters (any of them).

Only OUTGOING external bandwidth is counted.

All other traffic is free, including OVH to OVH transfers.

mks
01-11-2010, 10:31
Please confirm for me:
Inbound bandwidth - data coming into the server from the Internet
Internal bandwidth - data inbound/outbound from OVH server to another OVH server

Please confirm for me that those 2 types of traffic are not counted in the monthly bandwidth allotment and so they are considered free.

Myatu
23-10-2010, 13:19
mks, this was an idea Oles was contemplating... see http://forum.ovh.co.uk/showthread.php?t=4412

Given the lack of follow ups and a few outstanding questions, I don't think it has actually been implemented at this time.

mks
22-10-2010, 23:14
That's a reasonable price. I will utilize that next time my server goes down. Do they also have any resolution time guarantee for the incidents?

How do you open a paid incident ticket? I opened an incident ticket before and it took them over 24 hours to respond and only told me to look at the logs myself. Is that incident ticket different from a paid incident ticket?

YouWhat
22-10-2010, 20:11
Yes they do

The incident support now has two missions:

- If you have a problem with your service and you do not know where to look or what to look for, you create an incident ticket to request a diagnosis on your problem. This diagnosis will be charged €20euro ex VAT. If during the diagnostic OVH has solved the problem, you only pay €20euro ex VAT. If the problem is not fixed, OVH provides you with a diagnostic quote to solve the problem. If you like it, you can accept it and OVH does the job. OVH may also only provide estimates and in this case you do not pay for the diagnostic. If we figure out that we do not have the internal competances to do the job, you of course will not be charged for the diagnostic.

- If you have a problem with the service and you think that is the fault of OVH, you can create an incident in the same way as that of the charged diagnostic. OVH performs the diagnostic and if the problem is really on ths side of Ovh, all is covered by the service guarantee. However, if it is not the fault of OVH, we revert to the diagnostic principle and charge the €20euro ex VAT.

http://forum.ovh.co.uk/showthread.php?t=4217

- Quote from oles himeself (indirectly of course as he only speaks in french!!)

mks
22-10-2010, 20:00
Yes it has been reset now.

Does OVH have managed services for a fee?

I am not talking about the prepaid ones, but only pay to have something fixed.

Neil
22-10-2010, 13:35
Quote Originally Posted by mks
Bandwidth
The robot for resetting traffic starts early in the afternoon (in a few hours), so later today your bandwidth will be reset.

mks
22-10-2010, 13:25
Bandwidth

marks
22-10-2010, 13:17
sorry, what re-initialisation are you talking about?

YouWhat
22-10-2010, 12:44
Quote Originally Posted by mks
Does the reset occur at the beginning of the day or end of the day?
End of the day from past experience

mks
22-10-2010, 11:45
Does the reset occur at the beginning of the day or end of the day?

mks
22-10-2010, 06:27
Well they finally took the intervention off and I am now hit with the same DEFECT/END OF DEFECT problem. This problem has occurred on all the servers I have had with OVH.

yonatan
22-10-2010, 01:29
Use the support ticket interface and ask them about this..., while the server pings, but you cannot reboot - it might be that someone is working on your server currently...

mks
22-10-2010, 01:04
An intervention is being carried out on this server, at the moment you can't perform a hard reboot yourself

The server was not pingable then intervention took over. The server is now pingable but SSH is down. The server has been pingable since 2 hours but there is still intervention on it.

How can I get a status update on what is happening or did the support forgot to take off intervention? The weekend is coming up and I can't have this server down and unusable over the weekend. There is no mention of this current intervention under the intervention history.

yonatan
21-10-2010, 19:27
Quote Originally Posted by mks
That SSH is different, it does not boot from the server, we can't see what the problem is. What is the price rate to have an OVH tech look at the server? Many unmanaged providers do provide this extra service for a fee. I just need to know what the problem is so I can fix it on my own.

after you login to rescue pro you can mount your hard drives and check them manually .

that's the whole idea of the rescue pro interface, its a nice tool which you can use to look at your system ( software and hardware ) from "outside".

the only downside i see in the rescue-pro is an outdated version of LVM maybe someone in ovh can take a look at that for us ...


mks
21-10-2010, 18:21
Quote Originally Posted by Neil
Our servers are unmanaged, we cannot afford to provide the server at these prices if we were to investigate on each server when a customer says that is not working properly.
OK I understand that aspect.

Quote Originally Posted by Neil
No we will not look at the servers if you phoned, we provide all the tools to manage a server. If it is all your servers are going down, then I suspect a software problem maybe one of the applications you run is causing issues.

You have rescue mode which gives you SSH and has a hardware checker to rule out hardware issues.

Maybe some members of this forum will help you, but these kind of issues are out of OVH's support.
That SSH is different, it does not boot from the server, we can't see what the problem is. What is the price rate to have an OVH tech look at the server? Many unmanaged providers do provide this extra service for a fee. I just need to know what the problem is so I can fix it on my own.

Neil
21-10-2010, 14:24
Quote Originally Posted by mks
The server price is the same but it is not allowed even if I request the move after the next billing date.

I noticed the METERED and UNMETERED is on different IP blocks.

Is this a technical problem or is OVH just not allowing it?
We only allow Unmetered to Metered change when customers want better routing or have mistakenly ordered the wrong one, because of the Unmetered having low priority on on transit links.

Neil
21-10-2010, 12:55
Quote Originally Posted by mks
marks, my servers always have this problem of being pingable but not-bootable. Tech support tells me to check the logs but the logs indicate nothing. I was told that unless there is a verified hardware problem, no actual person will look at the server. Why is that?
Our servers are unmanaged, we cannot afford to provide the server at these prices if we were to investigate on each server when a customer says that is not working properly.

I had to keep going back and forth with the tech on guessing what the issue is when an actual person could've found out the issue in a few minutes just by looking at the server.

Even now, I still have no idea why my servers keep doing that. I understand these are unmanaged servers but you should at least get SSH up for the client, I can't manage it if I can't even get in SSH.

If I call instead of email, would that get a tech to look at the server?
No we will not look at the servers if you phoned, we provide all the tools to manage a server. If it is all your servers are going down, then I suspect a software problem maybe one of the applications you run is causing issues.

You have rescue mode which gives you SSH and has a hardware checker to rule out hardware issues.

Maybe some members of this forum will help you, but these kind of issues are out of OVH's support.

mks
21-10-2010, 12:39
marks, my servers always have this problem of being pingable but not-bootable. Tech support tells me to check the logs but the logs indicate nothing. I was told that unless there is a verified hardware problem, no actual person will look at the server. Why is that?

I had to keep going back and forth with the tech on guessing what the issue is when an actual person could've found out the issue in a few minutes just by looking at the server.

Even now, I still have no idea why my servers keep doing that. I understand these are unmanaged servers but you should at least get SSH up for the client, I can't manage it if I can't even get in SSH.

If I call instead of email, would that get a tech to look at the server?

mks
21-10-2010, 12:32
The server price is the same but it is not allowed even if I request the move after the next billing date.

I noticed the METERED and UNMETERED is on different IP blocks.

Is this a technical problem or is OVH just not allowing it?

marks
19-10-2010, 17:08
that could be caused for lots of different reasons, and you definitely can't tell just because of the emails.

So, the only thing that those emails mean is that the server stopped pinging for a period of time longer than what a normal reboot should take (our monitoring system doesn't trigger an intervention only with some seconds without pinging, it leaves some margin).

In normal cases, when the monitoring has detected a server down for a minimum time, it automatically programs an intervention (an engineers goes to the server to check). Hence the SLA for Level 1 interventions.

In your particular case, the server came back online by itself before the intervention was executed, so no more information. You should be able to get more info from the logs (maybe a fsck check made a reboot last longer than usual, ....)

mks
19-10-2010, 15:46
They say that the only way they can get somebody to look at the server is if they can confirm there is a hardware issue.

gueriLLaPunkMoB
18-10-2010, 01:20
Having issues here too:

http://forum.ovh.co.uk/showthread.php?t=4746

mks
16-10-2010, 19:06
I thought the server is not pingable if it's running fsck?

glidewave
16-10-2010, 18:11
While that is true (fsck), the issue was not fsck on my sp-storages as booting in vkvm, system got past that and stuck, and if left alone for days would just sit there; and since support was repetitive and cookie cutter "Software Diagnosis" - I had no desire to waste weeks of my time trying to get blood from granite, so each time reimaged (each time I could simply boot rescue test the drives and backup any/all data) - until the last time where I got fed up, reimaged and just returned the damn thing.

yonatan
16-10-2010, 16:48
Quote Originally Posted by Myatu
To expand on that: an fsck will take quite a long time...

When you do a hard reboot, it's quite possible an fsck is performed on the disks, hence none of your services are up until its done (the disks remain unmounted while the fsck is ongoing), but pingable.

You can disable fsck at boot time by changing the last digit in a row in /etc/fstab to zero (0). Example:

Before:
Code:
/dev/sda2       /               ext3     noatime,errors=remount-ro 0       1
After:
Code:
/dev/sda2       /               ext3     noatime,errors=remount-ro 0       0
Or simply avoid doing a hard reboot

fsck should be disabled only if you would keep a close look after the filesystem state.

If this is disabled.
It's recommended , to reboot into rescue mode, and manually run fsck in a case of any problems with the server,
in that way, you can monitor the progress and see if there are any issues.
once you are done running it manually , you can reboot your server and stay calm while you know the system is clean :-)

don't underestimate the power of fsck!

Myatu
16-10-2010, 15:45
Quote Originally Posted by yonatan
might be fsck ?
To expand on that: an fsck will take quite a long time...

When you do a hard reboot, it's quite possible an fsck is performed on the disks, hence none of your services are up until its done (the disks remain unmounted while the fsck is ongoing), but pingable.

You can disable fsck at boot time by changing the last digit in a row in /etc/fstab to zero (0). Example:

Before:
Code:
/dev/sda2       /               ext3     noatime,errors=remount-ro 0       1
After:
Code:
/dev/sda2       /               ext3     noatime,errors=remount-ro 0       0
Or simply avoid doing a hard reboot

Thelen
16-10-2010, 13:00
SP storage 09/10 has major issues, the new reloaded ones are fine. the old ones are becuase of the bad disks...

mks
16-10-2010, 01:30
This is quite interesting too, when you do a hard reboot through the manager, they actually have this as one of the reason which is the same problem I keep experiencing:
The server is responding to pings but ssh is down

Is this problem so common at OVH that they even put that as one of the reasons for rebooting?

mks
16-10-2010, 01:25
Quote Originally Posted by glidewave
a couple of servers we had did that frequently - dropped them since support was less than helpful.

A similar thing occurs on soft raided boxes which randomly fail to come back up - happened several times over a year and ended up dropping that sp storage this month.
Interesting, the type of the servers I have with them with this error is also the SP Storage. I have been trying to get support to hook up a monitor to see what's up but they would won't do it for me. That's the quickest way to find out what's wrong.

Another server went down today after I posted this thread. That's 2 different servers in a span of just hours. Support has not been helpful so far. If they would send a tech to look at the server, we can find out what the problem is instead of guessing back and forth but they wouldn't do that. Don't know why.

glidewave
15-10-2010, 22:36
a couple of servers we had did that frequently - dropped them since support was less than helpful.

A similar thing occurs on soft raided boxes which randomly fail to come back up - happened several times over a year and ended up dropping that sp storage this month.

Winit
15-10-2010, 20:28
Just you.

yonatan
15-10-2010, 16:34
might be fsck ?

mks
15-10-2010, 15:11
This has happened to 2 of my servers already.

I would get a DEFECT email that the server was unpingable and few minutes later get a END OF DEFECT email saying the server is now pingable and intervention is cancelled. But although the server is now pingable, SSH and all other services are down.

All hard drive, memory, and CPU test returns no errors also.
12 hours later, the services are up automatically without intervention.
The logs does not show any errors.

Does anybody else get this type of problem or is it just me?