You are viewing limited content. For full access, please sign in.

Question

Question

issues with connecting to Laserfiche server when SQL or File server are restarted

asked on April 1, 2014

 

We have a 3 server setup.

 

  1. Laserfiche Server, hosts the server, weblink & web access
  2. SQL server, hosts our SQL databases
  3. File Server, hosts our files for various servers

 

Whenever our file server or domain controller is restarted Laserfiche looses connection and users cannot login to the Laserfiche client.

A restart of the Laserfiche Server Service resolves the issue

 

whenever our SQL server is restarted Laserfiche looses connection and users receive a general 9008 error and cannot login to the Laserfiche client.

A restart of the Laserfiche Server Service DOES NOT resolve this, nor a restart of the SQL server service on the SQL server, a full restart of the Laserfiche Server did resolve the issue.

 

Is there something I am missing in the configuration on the Laserfiche server that would allow these other components to be restarted for updates/patches without losing connection on the Laserfiche Server?

0 0

Replies

replied on April 1, 2014

Is your domain controller also your file server?

 

What type of SQL authentication are you using?

 

The answer may be switching from Windows Authentication to SQL Server Authentication in your DBMS setup.

1 0
replied on April 1, 2014

We have two domain controllers separate from the file server

 

DC1

DC2

File Server

 

I'm not 100% sure which authentication we have configured. I know my users are using windows authentication to login to the Laserfiche client.

 

would that configuration be set in the Laserfiche admin console or do you mean on the DBMS DB setup specifically?

0 0
replied on April 1, 2014

In the admin admin console I'm showing;

 

driver: SQL native client

User: Sa

 

it's our SQL sa account

0 0
replied on April 1, 2014

Ok that is SQL Authentication. Is there any information in the event viewer that might give more details.

 

Some things I would look for when the 9008 error is occurring.

 

Can you use LF Client directly on the LF Server?

Can you access the repository from Admin Console?

Can you access the SQL engine from the LF Server?

1 0
replied on April 1, 2014

If you right click your repository in Admin Console and select Change DBMS Connection it will show you what type of authentication you are using to access the SQL Engine.

 

If it is set to Windows Authentication then I would setup a SQL server account and switch it to SQL Server Authentication. There may be a blip of downtime while you do this and verify the connection is up again.

 

I really can't explain how restarting a file server would effect your SQL authentication. However 9008 is a SQL error and obviously it is not a problem with the database since once the connection is back online everything is functional again.

 

Since the problem is not resolved by restarting the service but only by restarting the host OS all signs point to a Windows Authentication being conflicted with somehow.

 

There is no reason to take the users off Windows Authentication. That doesn't appear to be having any problems.

0 0
replied on April 1, 2014

I've not received this same error before.

 

The SQL error appears to have been after our SQL server was restarted.

 

We received a different error, such as "user is not licensed or all licenses are being used".

When that happened and I tested using Laserfiche authentication I was able to login to the repository. that was not the case with the SQL error.

0 0
replied on April 1, 2014

So here was the actual error from the event viewer. looks like because SQL dropped unexpectedly it must have thrown Laserfiche into a loop of trying to commit an action it couldn't.

 

Database statement execution error encountered. Session ID: ; Dialog ID: ; Repository: 'CNWInternal'; CLI routine: SysLogin; Message: Failed to retrieve error information!; SQL Statement: ''; SQL variable: ''; SQL State: .

0 0
replied on April 1, 2014

take that back, found the original at 5:52pm , found that my IT had restarted the server for sluggishness.

 

Database statement execution error encountered. Session ID: 2723; Dialog ID: 323118; Repository: 'CNWInternal'; CLI routine: SysActivityLogOpen; Message: [Microsoft][SQL Native Client]TCP Provider: An existing connection was forcibly closed by the remote host.

; SQL Statement: 'select top 400 asn, who, event, path1, path2, tocid1, tocid2, tocid3, uuid1, uuid2, uuid3, etype, old_obj_name, new_obj_name, old_obj_id, new_obj_id, prop_name, prop_id, flags, events from activity_log where asn <= :asn<bigint> order by asn desc'; SQL variable: ''; SQL State: 08S01.

0 0
replied on April 1, 2014

If I turn off my SQL server I get a Communication Link failure and a No process is on the other end of the pipe message in the event log from the LF service.

 

As soon as my SQL server comes back online I can once again use LF Client and Admin Console. It appears to function correctly. I am on version 9.1.1. What version are you on?

0 0
replied on April 1, 2014

It looks like I am on Laserfiche server 9.1.0.

 

I believe because the server was brought down unexpectedly someone may have been accessing the system and caused an error in the system when it was being brought down.

 

i'll run the updates to v9.1.1 as well

 

Is there anything in particular I should look at regarding the Domain controller issue? I've noticed that quite a few times on Monday when we come in and they were restarted sunday night for maintenance that Laserfiche doesn't automatically reconnect to allow for windows authentication?

0 0
replied on April 1, 2014

Do the users get the same 9008 error when this domain controller restart happens? Or do they get another problem?

0 0
replied on April 1, 2014

The users receive a different error when the DC is restarted, trying to find the specific wording but it's similar to;

 

"user is not licensed or all licenses are being used".

0 0
replied on April 1, 2014

Here was the details I had submitted to support on roughly 11/1 for the issue. they said an update to 9.1 should fix it.

 

error we are receiving "the user has not been allocated a named user license or the maximum number of sessions ha sbeen reached" [9030]. we have a process that monitors our windows patches/hotfixes and automatically updates the servers and restarts as needed to apply these updates.

this seems to happen over the weekend and this past weekend after one such restart when users tried to login monday morning this is the error they received. I login and verify all services are running and they are, to resolve the problem I restart the Laserfiche Server service and everything is then accessible

if I do a standard shut down or restart the server comes up correctly with services started and users able to access.

trying to troubleshoot the cuase of the access not being granted after the forced restart by our update application
 

0 0
replied on April 1, 2014

That is a very clear message and easy to verify.

 

In the administration console go to All Sessions and verify the user does not have more than 3 active sessions already open. (Each account has 4 max from what I understand).

 

Check the user's account and make sure they are set as a Full Named User.

 

Finally make sure you have not over-allocated your named users, this can sometimes happen even though the software attempts to prevent it.

 

If it turns out the user is assigned a license and you have enough licensing then you will want to contact support. That would be very unusual and unexpected behavior.

1 0
replied on April 1, 2014

you've got it, verified those items and contacted support.

 

IT was really quite odd while receiving it (also driving my network admin nuts because I'd get 30 emails at 7:45 Monday morning because we did patches sunday).

 

Thanks for all the feedback!

0 0
replied on May 15, 2014

We have a customer getting this same error now. 

 

Database statement execution error encountered.  Session ID: ; Dialog ID: ; Repository: 'DR05'; CLI routine: SysLogin; Message: Failed to retrieve error information!; SQL Statement: ''; SQL variable: ''; SQL State: .

 

I see other customers getting the statement execution error but with specific details instead of failed to retrieve error information. I think the problem is that we are not getting the critical information we need. I am not sure how to correct this. The only thing I can think of is running a packet trace on sql traffic from the LF service to the SQL Server. Thankfully they are on different machnes!

0 0
You are not allowed to follow up in this post.

Sign in to reply to this post.