Restore the Wazuh alerts from Wazuh 2.x

New in version 3.1.0.

After upgrading Wazuh from 2.x to 3.x, the old alerts will not be lost, however, they cannot be visualized in Kibana due to a change in the Wazuh alerts' template. In order to access the old alerts and visualize them along with the new ones, the indices need to be reindexed to apply the new mapping.

To do so, download the script (the script is available here) and Logstash's configuration file called restore_alerts.conf (the configuration file is available here):

# curl -so
# curl -so restore_alerts.conf

The Wazuh alerts can be restored in two different ways:

  • By restoring them from Elasticsearch to index them in Elasticsearch.

  • By restoring them from the Wazuh manager to index them in Elasticsearch.

Restoring the Wazuh alerts

  1. Stop Logstash:

    # systemctl stop logstash
  2. Run the script as superuser to reindex the old Wazuh alerts:

    # ./


The script needs Logstash to be installed on the same machine. If the script can not find Logstash, it will install Logstash to complete the task. After the reindexing has finished, Logstash can be uninstalled.

  1. Insert the requested parameters into the prompts from the script as defined below :

    • reindex_type: Indicates the reindexing type. It can be either set to ELS2ELS, for reindexing from Elasticsearch to Elasticsearch, or WM2ELS, for reindexing from the Wazuh manager to Elasticsearch.

    • elastic_ip: Is the Elasticsearch IP address. By default, set to localhost.

    • dateFrom: Start date as YYYY-MM-DD.

    • dateTo: End date as YYYY-MM-DD.


    If the user wants to reindex only a single day, the dateFrom and the dateTo should be set to the same date.

    The user can also execute the script and add the values for the parameters as arguments:

    # ./ date_from(yyyy-mm-dd) date_to(yyyy-mm-dd) elasticsearch_ip ELS2ELS|WM2ELS
  2. Once the script has finished, Logstash can be started again:

    # systemctl start logstash

Verifying the reindexing process

Once the reindexing of the alerts is completed, the user can confirm that the process was successful by querying the Elasticsearch indices:

# curl "http://localhost:9200/_cat/indices?v"

If the process was successful, the example output of the command looks as follows:

health status index                           uuid                   pri rep docs.count docs.deleted store.size
green open   wazuh-alerts-3.x-2017.12.12     vQ4YXsTuQLSDMnLk_Lp2Kw   5   1         58            0    115.1kb        115.1kb
green open   .kibana-6                       0jtvjQ4ERLmkKbCJ7Pl4Ww   1   1        241          110    226.5kb        226.5kb
green open   .wazuh-version                  AqVHhREjSgCpx07LJ45Dkg   5   1          1            0      7.1kb          7.1kb
green open   wazuh-alerts-2017.12.12         T3SZQRHGQEOBbVi79nDmhg   5   1         58            0    239.2kb        239.2kb
green open   .wazuh                          GV7tVKXsSb-BocyjxC07Iw   5   1          0            0      1.2kb          1.2kb