Unsolved
3 Posts
0
2440
Networker and AWS Glacier
I am using Networker 9.0.2 with cloudboost and a local datadomain. My current configuration is setup to backup to the datadomain and then clone to AWS. In this configuration only S3 is being used. I want to use Glacier. In what scenarios is AWS Glacier supported?
ble1
2 Intern
2 Intern
•
14.3K Posts
0
October 24th, 2018 06:00
I don't remember seeing Glacier mentioned (it would be idea for archiving loads via cloudboost), but I also didn't read yet 18.1 notes (I believe so far they used S3, right?)
captain_p
3 Posts
0
October 24th, 2018 07:00
So I have seen some documentation from Amazon talking about using their Virtual Tape Library Gateway for things like this and Networker is on the supported list but I haven't been able to find any documentation from EMC. I surely dont want to use a method thats not supported by EMC/Dell.
bbeckers1
2 Intern
2 Intern
•
137 Posts
0
October 25th, 2018 12:00
NW9.0.2? Or do you mean rather nw9.2.0? As far as I know nw9.0 only goes as far as nw9.0.1.9? and not into a 9.0.2 tree.
I assume you use an on premise DD while you clone towards a cloudboost instance running in AWS? so no cloud tier approach via the DD directly?
CompGuideApp states for nw9.2 cloudboost support:
"Virtustream Storage Cloud, EMC ECS Appliance, EMC ATMOS, OpenStack Swift, Microsoft Azure Storage, Amazon S3, Google Cloud Storage, AT&T Synaptic Storage, Scality."
So only S3 for Amazon.
be aware that cloudboost will drop the emc cloud portal approach used to manage cloudboost in June 2019 and with cloudboost 18.x will move to an integrated management portal (as customers did not like having cloudboost connecting to the internet). however this also appears to require moving to nw18 also. At the moment there is no target release for nw18. so depends if you'd want to upgrade to nw18 to be able to keep on handling cloudboost...
due to the newest DDVE 4 also having the option to have a object storage as primary tier (currently aure and aws only), we are looking into replacing cloudboost with ddve in a complete azure implementation (NW, clients and cloudboost all in Azure). so then both solutions would have a object storage backend. Due to this DDVE would also be somewhat cheaper than before as it will be using object storage (although DDVE still requires some (un)managed meta data disks). Also due to better expected dedupe ratio we're going to swap to DDVE instead of cloudboost.