This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

MS SQL DB Connection Broken in 23.4

I just upgraded a Filr environment to 23.4 and now tomcat can no longer connect to the external MS SQL DB.  I see errors like this in the Catalina.out:

com.microsoft.sqlserver.jdbc.SQLServerException: "encrypt" property is set to "false" and "trustServerCertificate" property is set to "false" but the driver could not establish a secure connection to SQL
Server by using Secure Sockets Layer (SSL) encryption: Error: Certificates do not conform to algorithm constraints.

The SQL connection was never encrypted and this has been the DB through many many versions of File. Did something change in 23.4? Any idea how to get it re-connected?

Matt

  • 0

    I see the same thing, what version of MS-SQL do you use? My customer is on 2016

    /Lennart

  • 0   in reply to 

    I have the issue duplicated in my lab.  It is specific to MS SQL versions 2016 and older.  The 23.4 documentation states MS SQL 2019 and 2022 are supported.  Versions after 2016 use SHA256.  Refer to this article from MS for more details:

    https://learn.microsoft.com/en-us/sql/database-engine/configure-windows/sql-server-and-client-encryption-summary?view=sql-server-ver16

    Check the catalina.out file and you will most likely see something like this:

    06-Nov-2023 23:10:50.565 WARNING [Catalina-utility-3] org.apache.naming.NamingContext.lookup Unexpected exception resolving reference
        com.microsoft.sqlserver.jdbc.SQLServerException: "encrypt" property is set to "false" and "trustServerCertificate" property is set to "false" but the driver could not establish a secure connection to SQL Server by using Secure Sockets Layer (SSL) encryption: Error: Certificates do not conform to algorithm constraints. ClientConnectionId:3f99e7ea-2ebe-4f12-8e88-dd559b4e6278
            at com.microsoft.sqlserver.jdbc.SQLServerConnection.terminate(SQLServerConnection.java:4026)

    I am working with engineering to come up with a workaround.  For now, the answer is to upgrade MS SQL.  I'll update this thread if we are able to come up with a workaround.

  • 0 in reply to   

    Hi,

    Yeah I noticed error and documentation, unfortunately first after upgrade.

    Customer got sql 2019/2022 in environment so they are migrating to that db-server.

    Question is if we have do anything else on Filr appliances to be able to use that server or if it will work automagically

    /Lennart

  • 0   in reply to   

    I saw that error in the log files and this site is on MS SQL 2016 SP3. But the SQL connection is not encrypted, so I don't understand what certificate is being talked about here?  What is it referring to exactly?

    The release notes do list 2019 and 2022, but don't really indicate the removal of support for previous versions. The install docs haven't been updated to remove SQL Server 2017 and 2016 SP1 either:

    https://www.microfocus.com/documentation/filr/filr-23.4/filr-inst/system-requirements.html

    Microsoft SQL

    • SQL Server 2022

    • SQL Server 2019

    • SQL Server 2017

    • SQL Server 2016 SP1

    That is in the 23.4 install docs.  That should be updated if it is not correct.

    Matt

  • 0 in reply to   

    Customer migrated the database to newer version, so now login works but all settings are gone.

    So that did not help really.

    I figure we will role back and migrate database first then after that upgrade Filr.

  • 0   in reply to 

    When the Filr node(s) are being updated, there is a script that runs which updates the database.  Since Filr was unable to connect to the database, the script failed.  If the connection to the database succeeds after the Filr node has been updated, you will need to run the DB update script manually.  Run the /var/adm/update-scripts/filr-db-5.0.0-150400.74.123.1-update-database.sh script.

  • 0 in reply to   

    Ok, I have done that now on one node, the script run without any error

    Trying to launch administration (not 9443) interface gives this error

    "Error
    No available node to select for read"

    users are able to login but no files are visible, both in home folder and net folders

    Any suggestions?

    /Lennart

  • 0   in reply to 

    Sounds like maybe there is an issue with the search appliances.  Are you able to log in to the 9443 interface on the search appliances?  Check System Services and make sure all services are in a running state.  Check the appserver log on the Filr node(s) and look for " Connection refused to host:" followed by the host name of your search appliance.  If you see that message check to make sure the search appliances are listening on port 1199.  If they are not, you can try running the search update script filrsearch-ssfclosed-indexer-lucene-5.0.0-150400.5.86.1-update-server-xml.sh located at /var/adm/update-scripts on each of the search appliances.  The other option is to roll back and try the update again as it is hard to know the current state of your Filr environment.  You also might consider opening a case.

  • 0 in reply to   

    filrsearch-ssfclosed-indexer-lucene-5.0.0-150400.5.86.1-update-server-xml.sh and reboot fixed my issues

    Thanks

    /Lennart

  • Suggested Answer

    0   in reply to 

    Thank you for the update.  Happy to hear it is working.