Good afternoon
We currently have a database connection called ENIQ, this does not index data in the indexer, it makes queries directly to the server.
For them there are many queries like the following:
example:
| dbxquery connection = ENIQ query = "SELECT CONVERT (char (23), DATETIME_ID, 121) as 'Date_hour', round (sum (ggsnDownlinkBytes * 😎 / (900 * 1000 * 1000), 2) as Downlink_Total, round (sum ( ggsnUplinkBytes * 😎 / (900 * 1000 * 1000), 2) as Uplink_Total from DC_E_GGSN_GGSN_RAW where datepart (dy, DATETIME_ID)> datepart (dy, getdate ()) - 3 and (GGSN = 'SAGG07' or GGSN = 'SAGG08' )
Reviewing DBX - Connection Health - Connections, we have a total of 10,000 to 20,000 daily connections of this type as there are panels that are constantly monitored and re searches refresh every 5 to 15 minutes.
They inform us that from the source server, the connections are not closed, they are stuck and this has caused multiple drops in the database server.
According to the official documentation http://docs.splunk.com/Documentation/DBX/3.1.1/DeployDBX/Troubleshooting
it is required to modify the db_connections.conf, the following parameters were added to the connection:
maxConnLifetimeMillis = 1800000
maxTotalConn = 8
but it seems that these configurations will only be applied to dbquerys stored using cron, and these queries are in many dashboard querying directly since they have been defined in that way without using a collect to not consume license.
Is there any recommendation to force the closure of these dbquerys ?, this connection is using version 16 sybase and connector version 7
Evidence of results will be attached per day of ENIQ connection and details of connections from source.
Hi,
I am also facing the problem with the database=hive2. creating a lot of connections.
Kindly have any suggestions on it.
Hi.
According to the documentation delivered by splunk support, http://docs.splunk.com/Documentation/DBX/3.1.1/DeployDBX/Troubleshooting, does not apply to my case, because the connections to the database were dbxquerys and they are not can be administered by the splunk_db_connect app. Review the official account if there is any parameter that will help you in your problem:
http://docs.splunk.com/Documentation/DBX/3.1.3/DeployDBX/database_typesspec