korsikov - Thanks for the post.
Can you be sure that the service account running the reload (i.e. running the Qlik services) is a root admin and that sa_repository is also a root admin on the server?
We have seen this before and sometimes adding rootadmin role to any account that is reloading the apps can resolve it.
Let me know. Thanks!
Thanks for the quick response. But he has already come to the same decision.
For the first time confronted with the fact that we have to do on behalf of the user running the Server service which root administrator.
There are other servers that are upgraded from older versions of the server in these services user does not have the right root admin but reload performed normally. Why can this be?
Can you help us debug this by running the Qlik Repository service with the -record parameter?
You do this by changing the option in the windows Services for Qlik Repository Service and adding "-record" as a startup parameter and then restarting the service (which will restart all services).
Try reloading both apps a few times and then gather the logs which are produced.
let's be clear.
I do it on the server where faced with this problem.
1. add to Qlik repository service option -record
old record "C:\Program Files\Qlik\Sense\Repository\Repository.exe" -iscentral change to "C:\Program Files\Qlik\Sense\Repository\Repository.exe" -iscentral -record
Restart Qlik Repository service
Delete role Root admin from qlik services user (qvsuser) then try reload application and then gather the logs
It's simpler than that I believe.
- Go to "Services"
- Stop the Qlik Repository Service
- In the "Parameter" input box, add -record
- Start the Qlik Repository Service
- Start the other Qlik services
- Restart the Qlik Service Dispatcher
Start the License Monitor reload and the Operations Monitor reload.
When they finish, check the repository logs here:
If you can, email them to me firstname.lastname@example.org so the developers can check them.
Okay - the way to do this is to:
- Stop the Qlik Sense Repository Database (stops all other services)
- Stop the Qlik Sense Repository Service (if not already stopped)
- In the "Properties" of the Qlik Sense Repository Database, add -record to the parameter at the bottom.
- Click Start
- Repeat 3 and 4 for the Qlik Sense Repository Service
- Start all the other Qlik Sense services
- You may have to restart the Qlik Sense Service Dispatcher at the end
- Reload both Monitoring Apps (see the failure)
- Send email@example.com the following two logs, which are found in c:\
- Turn off qrs recording by restarting all the services.
currently having the same problem, just posted on a separate thread. here is the snippet of the error from the load script of my license monitor. I made the sa scheduler and sa repository root admin but that didn't resolve it.
2016-07-22 08:26:52 0457 CALL qrs_userAccess
2016-07-22 08:26:52 0458
2016-07-22 08:26:52 0478 *** CONNECT TO ***
2016-07-22 08:26:53 0480 RestConectorMasterTable:
2016-07-22 08:26:53 0481 SQL SELECT
2016-07-22 08:26:53 0482 "id" AS "id_u0",
2016-07-22 08:26:53 0483 "createdDate",
2016-07-22 08:26:53 0484 "modifiedDate",
2016-07-22 08:26:53 0485 "modifiedByUserName",
2016-07-22 08:26:53 0486 "lastUsed",
2016-07-22 08:26:53 0487 "quarantined",
2016-07-22 08:26:53 0488 "quarantineEnd",
2016-07-22 08:26:53 0489 "__KEY_root",
2016-07-22 08:26:53 0490 (SELECT
2016-07-22 08:26:53 0491 "id",
2016-07-22 08:26:53 0492 "userId",
2016-07-22 08:26:53 0493 "userDirectory",
2016-07-22 08:26:53 0494 "name",
2016-07-22 08:26:53 0495 "__FK_user"
2016-07-22 08:26:53 0496 FROM "user" FK "__FK_user")
2016-07-22 08:26:53 0497 FROM JSON (wrap on) "root" PK "__KEY_root"
2016-07-22 08:26:53 14 fields found: id, userId, userDirectory, name, __FK_user, id_u0, createdDate, modifiedDate, modifiedByUserName, lastUsed, quarantined, quarantineEnd, __KEY_root, __extra_,
2016-07-22 08:26:53 Error: QVX_UNEXPECTED_END_OF_DATA: HTTP protocol error 403 (Forbidden):
2016-07-22 08:26:53 The server refused to fulfill the request.
2016-07-22 08:26:53 Execution Failed
2016-07-22 08:26:53 Execution finished.
Thanks! Yes, this is part of the mystery. Some reloads fail when user account does not have RootAdmin rights.
But with the other non-RootAdmin system - be sure to check that you get User data back (UserName). We have seen that the reload is successful without RootAdmin rights but there is no User data fetched from the QRS.
Coming late to this discussion, but having the same problem.
Trying to reload the Operations Monitor that ships with Sense Enterprise 3.0.1, and get this error:
The following error occurred:
Could not fetch connection by name using identifier qrs_user
The error occurred here:
LIB CONNECT TO 'qrs_user'
Have made the sa_repository user RootAdmin. Btw, I didn't see any reference to sa_repository user above, but that would be the user - right? AFAIK that name is configured by the Sense installer, which makes it a bit strange if it is called different things on different systems...
Running the reload while RDP:d into the server, logged in as a Windows user that is both domain Admin in Windows and RootAdmin in Sense.
Virtual proxies were mentioned earlier - we do have a bunch of them.
I seem to be having the same problem.
However, using QMC, i am able to load the license monitor, but not the operations monitor.
(View the .log file for the reloading of operations monitor as attached !.)
I am using virtual proxy with one node.
I have appended the prefix of the virtual proxy to the URL of the data connections (qrs_*) but to no avail.
I have set the reload user as root admin and these includes Sa repository and Sa scheduler.
Do note that i have read and tried these solutions.
Monitoring a Qlik Sense site ‒ Qlik Sense (Updating data connections due to prefix of virtual proxy).
Setting reload user, sa repository and scheduler to root admin.
Import and replace operations monitor on the monitoring stream.
Thanks all, appreciate your help.
Hi gameoverzx -
Thanks for posting after reading the help instruction and trying that out.
It appears that you are connecting fine to the QSR database with the REST connector but you are not fetching any data - which means that the data connections are set up in a way to connect okay.
This leaves at least two possibilities:
- No permission to view data.
- You say the service account running the Qlik Sense Engine and Qlik Sense Repository services has been made RootAdmin?
- If you set yourself as owner of the data connection qrs_app (and you are RootAdmin), go out to hub, create a new app, open data load editor, and click the add data from the qrs_app data connection. Do you see data there? (It should show more than just "Root" in the data preview drop down.)
- Something else related to your system .
- Is this an upgraded environment?
- Can you open pgadmin (program files\qlik\sense\repository\postgresql\9.3\bin) and connect to the QSR database? (localhost, port 4432, postgres maintenance db)
Thanks for your quick response.
1. No permission to view data.
1. Yes. Service account running is RootAdmin and Qlik Sense Repository services has been made RootAdmin as well. (sa_scheduler and sa_repository).
2. I have obtained the following screenshots for following the steps provided.
Does this mean that as a RootAdmin, i do not have permission to view data.
2. Something else related to system.
1. To my current knowledge, this is not an upgraded environment.
2. I am unable to locate the bin file within the folder. (Attached below)
Regarding REST data connection - it looks like you are not finding data at the endpoint. Is this multi-node? If so, update the REST data connections' connection string URL to replace "localhost" with the fully-qualified domain name of the central node (or wherever the QRS database resides).
Also, this might be the issue with the ArchivedLogs Folder.
If you are running multi-node with sync, change the ARchivedLogsFolder data connection (in QMC) to point to the central node instead of to c:\programdata...
Or if you have shared persistence, use the ArchivedLogsfolder path listed in the "ServiceCluster" section in QMC.
Regarding the pgadmin, it is under Program Files, not ProgramData. Sorry, that wasn't clear in my original email.
This is a single node.
I have shared persistence. Have changed the ArchivedLogsFolder data connection path to the folder listed in the service cluster.
Still unable to reload operation monitor successfully. Thanks for your help, I think that this has to do with the connection string rather than the user rights to reload.
Yes, something sounds strange here. I would try creating a new REST data connection to the app endpoint using https://FQDN_where_QSR_db_is/[any virtual proxy prefix with windows authentication/qrs/app/full
Check the existing qrs data connection to get an idea of the other settings.
If you can't access any app data (as root admin) with this new data connection, then there is something more going on.
I would like to add on just in case i am off track. The initial problem was with reloading the operations monitor AND the license monitor in qmc and this is because the data connection "ArchivedLogsFolder" is unable to locate the file location. I have checked and there is no file labeled "ArchivedLogs" within C:\ProgramData\Qlik\Sense\Repository
Since then, i have changed the connection string of the ArchivedLogsFolder to C:\ProgramData\Qlik\Sense\Log\
This allowed for the reloading of the license monitor but not the operations monitor.
ARchivedLogs folder is not under Qlik\Sense\Log but is either under \\FQDN_of_central_node\c$\ProgramData\Qlik\Sense\Repository\Archived Logs (in the case of single node or multi-node sync) OR at the path of the Service Cluster's archived log entries as shown in the "ServiceCluster" section of QMC (for shared persistence).
I have updated the QRS connection string from https to http , added Archive folder to program data ,my account is root admin, I can restart the qlik services,It was working before upgraded to 3.2.4, same account I am using now, Please help me to make monitoring apps working again.
Thanks in advance
rohitkumar1609 - changing the URL to http will only work if your proxy is set to allow HTTP (check the proxy settings in QMC).
And you need to be sure you have a virtual proxy with "Windows Authentication" which either does not have a "virtual proxy prefix" or, if it does have a prefix, the qrs_ data connections' URL has been updated to include that prefix (https://myserver/prefix/qrs/app/full for example).
When you say "Monitoring App" is still not working, are you referring to the Operations Monitor app is not reloading successfully? (Or the app doesn't look right when you open it even after successful reload?)
ArchivedLogs folder should look like this:
\Server2 (if you have multiple nodes!)
This is a Qlik-supported app, so you can submit a help ticket.
What they (and I) would ask for are:
- Version of Sense
- script logs from the latest reload of the apps (in ProgramData\Qlik\Sense\Log\Script)
- single vs multi-node
- Any virtual proxies or changes there?
- Version of sense
- prior version of sense
- Shared persistence? (and location of ClusterShare)
Hi Tyler, I am sharing required details below:
1. Version : 3.2.4
2. Error in log:
FROM JSON (wrap on) "root" PK "__KEY_root"
2017-06-12 20:54:43 18 fields found: userId, userDirectory, __FK_owner, id_u1, __KEY_app, __FK_app, id_u2, createdDate, modifiedDate, modifiedByUserName, description, objectType, publishTime_u0, published_u0, approved, name_u2, __KEY_root, __extra_,
2017-06-12 20:54:43 Error: QVX_UNEXPECTED_END_OF_DATA: DNS error
2017-06-12 20:54:43 Execution Failed
3. Single node
4. No virual proxy
5.Prior version 3.2.2
6. No shared persistence.
Please let me know if you need more details.
Below path is mentioned in QRS_APP connection:
CUSTOM CONNECT TO "provider=QvRestConnector.exe;url=http://localhost/qrs/app/full;timeout=180;method=GET;autoDetectResponseType=true;keyGenerationStrategy=0;useWindowsAuthentication=true;forceAuthenticationType=false;useCertificate=No;certificateStoreLocation=CurrentUser;certificateStoreName=My;queryParameters=xrfkey%20000000000000000;queryHeaders=X-Qlik-XrfKey%20000000000000000%1User-Agent%2Windows;PaginationType=None;IsOffsetStartFieldHeader=false;IsOffsetCountFieldHeader=false;IsOffsetTotalPathHeader=false;IsNextTokenFieldNameHeader=false;IsNextTokenFieldPathHeader=false;IsNextUrlFieldPathHeader=false;"
I am trying to run IP flush command , will update you
Weird. So the server that is reloading the Operations Monitor is not resolving the "localhost" DNS.
I confess this is a bit out of my expertise (I would tell you DNS = donuts )
- No permission to view data.