@@ -78,10 +78,10 @@ Furthermore, the log analysis infrastructure based on Elasticsearch-Logstash-Kib
\section{Disaster Recovery}
On November 9th CNAF T1 Datacenter was flooded following a problem with a water pipeline, filling up to one/two rack unit.
On November 9th CNAF T1 datacenter was flooded following a problem with a water pipeline, filling up to one/two rack unit.
Because of this disaster, it was necessary to physically move our infrastructure to the INAF (CNR Bologna) datacenter in order to restore our services while maintaining the same network.
Following this event, the GARR team provide us a disaster recovery (DR) site with dedicated hardware in ReCaS-Bari datacenter. On such hardware a virtualization infrastructure based on VMware was prepared on which to operate autonomously.
Following this event, the GARR team provide us a disaster recovery (DR) site with dedicated hardware in ReCaS-Bari datacenter. On such hardware a virtualization infrastructure based on VMware was prepared on which to operate autonomously.
Our database systems (Oracle, MongoDB, etc) are replicated on disaster recovery sites using their own replication system (e.g. Oracle Dataguard). For database systems in which there was already a slave on the LNF site, an additional slave was added to the Bari site.
At the same time, database backups (e.g. made with Oracle RMAN) and a cold backup of INFN ERP system were also sent to the Bari Site.
