We have noticed an issue in my Splunk environment:
Data is getting duplicated twice in indexers. If i do a search in search head, the same events are coming in twice. this issue started today, earlier there is no issue with the data.
1) Checked the application logs whether same log is existing twice. Answer: No
2) Checked whether this issue is happening to one sourcetype OR only for one index. Answer: No it is affecting all indexers data.
Any other reason why this is happening? And what are the steps needed to prevent it?
Thanks in advance.
In case of duplicate issues, we need to check the following:
The following endpoint lists all files known to the tailing processor along with their status (read, ignored, blacklisted, etc...)
If you can not able to rectify the issue in the above scenarios, you can enable the DEBUG level using the following components.
To check if the events are duplicated, you can use follwoing SPL,
| eval md=md5(_raw) | stats count by md | where count > 1
For more information, kindly check, community: Troubleshooting Monitor Inputs
You have mentioned that all your data is getting duplicated, this sounds like a misconfigured outputs.conf
Can you confirm how your outputs.conf is configured?
Here's an example with 2 indexers which are in an indexer cluster named indexer 1 and 2, indexer acknowledgement is also turned on, SSL is not in use in this example:
defaultGroup = allIndexers
disabled = false
autoLB = true
useACK = true
#DO NOT EDIT THIS FILE! #Changes to default files will be lost on update and are difficult to
#manage and support.
#Please make any changes to system defaults by overriding them in
#apps or $SPLUNK_HOME/etc/system/local
#(See "Configuration file precedence" in the web documentation).
#To override a specific setting, copy the name of the stanza and #setting to the file where you wish to override it.
[tcpout] maxQueueSize = auto forwardedindex.0.whitelist = .* forwardedindex.1.blacklist = _.* forwardedindex.2.whitelist = (_audit|_internal|_introspection|_telemetry) forwardedindex.filter.disable = false indexAndForward = false autoLBFrequency = 30 blockOnCloning = true compressed = false disabled = false dropClonedEventsOnQueueFull = 5 dropEventsOnQueueFull = -1 heartbeatFrequency = 30 maxFailuresPerInterval = 2 secsInFailureInterval = 1 maxConnectionsPerIndexer = 2 forceTimebasedAutoLB = false sendCookedData = true connectionTimeout = 20 readTimeout = 300 writeTimeout = 300 tcpSendBufSz = 0 ackTimeoutOnShutdown = 30 useACK = false blockWarnThreshold = 100 sslQuietShutdown = false [syslog] type = udp priority = <13> dropEventsOnQueueFull = -1 maxEventSize = 1024