MSSQL11 Up and Down

dman

dman

Perch
Hello,

I'm seeing MSSQL11 errors periodically today for about 10 minutes at a time. This appears to have started yesterday morning. It happened 3 times yesterday and 4 times today. It looks like one just recovered at 7:02 PST. I checked the Network and Outages forum but didn't find anything related or recent. Any idea what's happening?
 
MSSQL 11 was not down at all. Please check the server up-time below.

C:\>net stats serve | find "Statistics"
Server Statistics for \\CL1-MSSQL11N
Statistics since 4/17/2013 10:17 PM

It might be network consumption issue at server end but not down for sure. If you contact us immediately on Live Chat at that time then it would be easy for us to help you instantly , but now I would like to request you to open a support ticket with mention the error message which you found. So we can investigate it further.
 
There have been a few attacks on sites that we resolve pretty quickly, but I have to say honestly I dont normally look to see what SQL server they link back to unless we get outage reports that the test query failed.

We aren't having any server side issue on that, and not network either, but it is possible it got slow if one of those attacks that got blocked was a site that used it and had high load due to it. The time frames would be similar from what you are saying to when I was working it pretty heavily.
 
Hey Stephen and Shubham,

I am continuing to see these errors periodically starting last Sunday, 11/3. Some of these errors only occur for a very short period of time because they are not triggering a monitor alert, which will check for the error again after 3-5 minutes. But, it is triggering an error alert from the custom web site error checking. I'll try to contact Jodo via chat when it occurs but often when I try to verify the issue it has resolved itself.

Maybe the NIC card is flapping? I've been getting a whole lot of these short-term errors Saturday and now it is continuing this morning. Below are the times these occurred on Saturday.

11/09
12:24 AM - 10:17 AM
9:52 PM -11:53 PM

I have seen the errors myself when the web site was unable to access the DB. I use a web site monitor that checks several sites every minute, couple of minutes or every 10 minutes after an error. The sites with errors are all using MSSQL11 and the sites that are not using a DB are not having errors. The errors I'm getting are below. I've seen these periodically in the past but since last Sunday they are occurring quite a lot...

  • [DBNETLIB][ConnectionOpen (SECDoClientHandshake()).]SSL Security error. - Microsoft OLE DB Provider for SQL Server
  • [DBNETLIB][ConnectionOpen (Connect()).]SQL Server does not exist or access denied. - Microsoft OLE DB Provider for SQL Server
  • [DBNETLIB][ConnectionOpen (PreLoginHandshake()).]General network error. Check your network documentation. - Microsoft OLE DB Provider for SQL Server
  • [DBNETLIB][ConnectionRead (recv()).]General network error. Check your network documentation.
  • Operation is not allowed when the object is closed. ADODB.Recordset
  • Timeout expired- Microsoft OLE DB Provider for SQL Server
Any idea why this is happening? Thanks for any help!
 
Can you give me a ticket number that gives the domain seeing this, or a private message to me that has the domain in question. I will monitor this through my shift and check details between the servers
 
We know the issue here, you are using the pass through external port for internal site, and this is the issue. Pass through is designed for external management convenience ONLY. Not for being used as the connection string to internal server to DB communications. This is why you are seeing issues, we don't monitor that for anything but up, as it isn't supposed to be used for any server to server communications.
 
Hey Stephen,

Thanks for your input as well! I understand I was using the wrong IP and port but it had worked for at least a year until 11/3. When I switched from 2000 to 2005 last November I could swear that was the IP and port I was given to use. I switched out the IP as suggested in my ticket and the errors have decreased a great deal but there have still be a few. I'll keep an eye on it and check back if the occurrence increases.
 
dman said:
Hey Stephen,

Thanks for your input as well! I understand I was using the wrong IP and port but it had worked for at least a year until 11/3. When I switched from 2000 to 2005 last November I could swear that was the IP and port I was given to use. I switched out the IP as suggested in my ticket and the errors have decreased a great deal but there have still be a few. I'll keep an eye on it and check back if the occurrence increases.
Click to expand...

We only directed people to use that for external access, never for internal. The reaosn is very simple why it would work and give errors sometimes, its mega masses of requests coming in, and adding web server traffic to it maxes out its capacity, which is exactly WHY we don't want web server traffic hitting it, it hurts everyone trying to get outside access. It is essentially a proxy on one server for all the servers, it works well for legit management requests, but never designed for continued high level back and forth queries. SQL can easily use 100mb/s b directional traffic on a busy site, and this x40 servers, would/will/can get out of hand.
 
You must log in or register to reply here.
Back
Top