Search the Community
Showing results for tags 'windows server'.
Found 1 result
Looking for help with a new implementation
Greg Mitchell posted a topic in DeploymentI have a very small VMWare Essentials environment. 2 Dell R340 Hosts, 1 socket each Backing up 3 Windows Server VMs running in VMWare esxi 7 u3 1 Linux vm lab machine ~3Tb footprint 2 Synology 4bay NAS RackStation RS820RP+, 18Tb storage each, 36Tb total storage I recently had a real-life disaster scenario. One of my older Synology NAS Devices lost two drives out of 4 in less than 24 hours. A real-life lesson that a RAID Array (or any redundant disk storage) is not a backup. It is a slightly more reliable place to store data that needs to be backed up properly. Also, I had not been backing up properly. One of the Windows VMs services a ~2Tb vhd, mostly Video, Pictures, and Office Documents. This drive is trouble to back up. It averages about 3gb overall block change each day, an average of < 50Mb a day in new bits. Backing this up on the block-level using the VMWare backend takes forever. By the time you compress it, dedup, and write it, the Full backup took 6-8 hours (again, not using the Nokivo tools). The daily incrementals only took 6 minutes. Looking at the full backups, I thought "that is slow, but not horrible." Restoring that file server took more than five full days, processing on brand new hardware, at ~30mb/s. During that time we were able to function as a company, but the pain was tremendous. I could restore individual files/folders upon request, but I could not launch that server until all of the blocks were in place. Needless to say, I am new to Nakivo because I am looking for a better tool. I would like to avoid disasters in the future. I have only just installed Nakivo Backup on my Synology devices I think it makes sense to create a Primary backup job that covers my 3 servers. I would eliminate that one vhd from the primary backup job and back those blocks up at the file level from the Windows OS to a secondary target. I could then replicate the vhd to the other host for a live backup, and process a compressed, deduped daily that gets tucked away, safe from ransomware. Does that make sense, or am I missing something? Additional note: To save money initially and because I was planning to use the free version of ESXi, I only bought one core per host. That makes fitting all of my active VMs on one host pretty hairy, and that is why I don't run a live replica of all 3 servers. If anyone would like to cure my ignorance, I am all ears. Thank you for your time. Greg