In this blog post we will be explaining how we are changing the iSCSI LoginTimeout and the reasons why we are doing it in this method.
Background Recently we were going through a code upgrade on our storage arrays and one of the recommendations from Pure support is to increase the iSCSI LoginTimeout prior to upgrading the code. For the uninitiated, VMware defines the LoginTimeout as “Time in seconds the initiator will wait for the login response to finish.
This blog explains how we configured and use Grafana to graph and visualize PDU data stored in InfluxDB.
Background InfluxDB is a time series database that we use to store dozens of metrics from all of our power distribution units (PDUs). This database is updated periodically with a python script that parses snmp data from our PDUs and stores it in the correct table and row. After getting the data into the db, we used Grafana to create graphs of some of the power consumption metrics so that we could easily see how much power was being used by each PDU.
Background VMware Affinity rules can solve many issues in a virtualized environment with out the need to create different clusters or datacenters to separate Virtual Machines. The specific need that we at Encore had for this was that when a vm is being cloned, if the VM was migrated to a different Host due to a vMotion, the custom spec with Kickstart files or Autounattended Syspreps would fail.
This was a random error that would only happened when vCenter or a host was overloaded.
