#VMUG ## Presentees - Dominik Schäfer - Steffen Matthias - DuncanYB ## Topics ### Einrichtung ESA - Schnelle Einrichtung des ESA Clusters - 2-Site mirrored - eigene Policy mit 2 sites aufbauen - HIstory of ESA ### ESA Overview - differences to OSA - NVMe only - no dedicated cache - no disk groups - higher performance - Requirements - vSAN ESA ready nodes - new capabilities - Log structured File system - reduced I/O - depending on I/O - small I/Os are collected - high I/Os are written instantly - available in 8.0U2 - compressing is done directly under VM - same as encyption - Latency - verly low, on a demo in 95th perc. under 0,00 ms - HCI Bench was used - Adaptive Raid-5 - 2 Schemes - 4+1 scheme(5 hosts+) - 1,25x capacity for resilient data storage - 2+1 scheme (1-5) - 1,5x cap. - Switchover during maintenance, outage happens automatically for the VMs - Auto policy - in vSAN 8U1 - is giving alerts in case of changes to cluster - showing up in SH - Gives recommendations fe. from R1 to R5 when new hosts are added - Can be split, vms can be migrated in waves to new Storage policy - 50G LACP worse than 25G single - article from dell - vSAN MAX/ HCI - vSAN from other Cluster can be used more easily - like any othe external storage - Clusters can be diferentiated between vSANMAX/HCI and COmpute Clusters - Compute Clusters do not need drives, only VMks with vSAN Service configured - No license is required for compute cluster(like external storage) - Snapshots - much lower latency on layered snapshots - in OSA files/blocks were moved on storage causing latency - in ESA only pointers are moved, much lower or no additianol latency ### Ransomware - VMware live Recovery - analyzes snapshots on entropy and other factors to discover Ransomware attacks - integrated with Carbon Black - Schedule of Retention - additional gui which creates snapshots and sends them to AWS/other external cloud providers