Virtualization Redemption?
-
No. We live in hurricane ally. an all out local outage is more likely 4 months of the year than a single server outage
-
@hubtechagain said:
not sure if it works with HyperV or if i'll have to find another local storage backup option to push to the NAS.
Is this how you are handling local failover? If the one node dies, you spin VMs up off of the NAS, using the NAS as temporary primarily storage?
-
@hubtechagain said:
No. We live in hurricane ally. an all out local outage is more likely 4 months of the year than a single server outage
Oh okay, I guess that makes sense.
-
@scottalanmiller said:
@Dashrender said:
@scottalanmiller Would StorageCraft in this case basically just sync his VM's from one client to another at a given point in time? I think that is real goal here. If the main site dies, Hub wants to connect to the VM host in the remote DC and just turn on the VMs there so they company is back on line.
Yes, it would be an incremental backup from one site to the other. As well as taking normal "on site" backups. It's a full backup utility.
That's correct. This design is especially useful as a warm failover when replicating between sites rather than replication between two hosts at the same site. As Scott points out, Hyper-V will replicate between hosts. VMware does this with vMotion as well. These provide native replication between two hosts on the same hypervisor platform.
Replicated backup image files become more useful when you:
A) you want to lower costs. For example, your hypervisor of choice charges for this feature.
B) you want to reach back into the past. For example, you get a virus and want the far side to use a virtual disk based on a point in time prior to the infection.
C) you have limited resources at the offsite. For example, your offsite is primarily a storage repository but you have some compute resources available. Those resources can be used as a temporary failover for onsite systems by implementing a backup image in the offsite storage. Public cloud services are a good example of this where you use offsite mostly for storage and only spin up a VM when needed. -
using the hyper v replication, how do the VMs behave? they're offline, replicating, then boom, tree crashes through building and smokes our server rack. i just remote into the DR server and spin em up?
-
@scottalanmiller said:
Paging @Steven
Yep... was doing a webinar with Redmond Channel Partners at the time so I responded as soon as I could.
Thanks for the ping!
-
@hubtechagain said:
using the hyper v replication, how do the VMs behave? they're offline, replicating, then boom, tree crashes through building and smokes our server rack. i just remote into the DR server and spin em up?
Basically, yes.
-
@hubtechagain said:
I'm currently using thinware for my local backups. not sure if it works with HyperV or if i'll have to find another local storage backup option to push to the NAS.
Veeam free could handle this. since you can just do a new full backup nightly or something.
Setup the powershell script and schedule it with task manager.
-
@hubtechagain said:
Well, were i to "network raid" my two hosts....I would not have enough space to handle the workload. Pretty sure that separate hosts is the way we need to stay at this client specifically. @Steven sadly you were a bit slow to the game and I think that the boys have me figured out
I'm currently using thinware for my local backups. not sure if it works with HyperV or if i'll have to find another local storage backup option to push to the NAS.I hear ya. Thanks for letting me join (late) and good luck!
-
@hubtechagain said:
using the hyper v replication, how do the VMs behave? they're offline, replicating, then boom, tree crashes through building and smokes our server rack. i just remote into the DR server and spin em up?
Yup, that basically sums it up. It's basically a "zero to five minute copy" sitting over at the other site.
-
@hubtechagain said:
using the hyper v replication, how do the VMs behave? they're offline, replicating, then boom, tree crashes through building and smokes our server rack. i just remote into the DR server and spin em up?
As I understand it, writes to one VM are also sent to the replica. The data sent happens at set intervals (e.g. 30s, 5min, and 15min). This creates redundant VMs on the two hosts with the replica slightly behind the source. When a tree crashes through the roof and craters one VM then the other simply keeps running (either it's the source and the target no longer receives updates, or it's the target and the host doesn't care).
-
You say that you don't have enough storage to run both workloads on a single server... is your DR site going to have enough storage and RAM for that?
-
@dafyre said:
You say that you don't have enough storage to run both workloads on a single server... is your DR site going to have enough storage and RAM for that?
Good point, we have to assume that both will failover at once, not just one at a time.
-
yes, i said above i'm adding 2 600gb drives and switching from raid 10 to 6. it's a lot to read, but i respect that. each server has more than enough processing power to handle all jobs, just unnderbought storage.
-
@hubtechagain Ooops on the underbuying storage, lol.
And yeah... Reading a lot of threads every day it hard to keep them straight, lol.
Out of curiosity... Why not switch local servers to RAID 6? ... aside from the headache of having to do it.
-
@scottalanmiller
Thank you for mentioning StarWind here! I really appreciate it!@hubtechagain
So, just want to confirm that StarWind Virtual SAN can do the job here. I would recommend you to install on all the 3 nodes that you have and run the 3-way replication on the storage level. Thus there will be no need to have redundancy on the RAID level, so you can use RAID 0 and have good speed and more disk space, while StarWind will do the fault tolerance on the storage level.As about DR considerations, I would recommend to take a closer look at Veeam Backup and Replication solution - they look like a good fit here.
Maybe I could give more precise advices if we would jumped on the quick call. Email me please when you will be available ([email protected])
-
@hubtechagain said:
yes, i said above i'm adding 2 600gb drives and switching from raid 10 to 6. it's a lot to read, but i respect that. each server has more than enough processing power to handle all jobs, just unnderbought storage.
I wouldn't say you under bought on storage. It's a DR situation. Assuming the business is OK with the performance at that level, I think this is perfectly acceptable.
-
@Dashrender said:
@hubtechagain said:
yes, i said above i'm adding 2 600gb drives and switching from raid 10 to 6. it's a lot to read, but i respect that. each server has more than enough processing power to handle all jobs, just unnderbought storage.
I wouldn't say you under bought on storage. It's a DR situation. Assuming the business is OK with the performance at that level, I think this is perfectly acceptable.
Well, he apparently underbought by two drives. But not a big deal. Adding them now, problem solved. I would assume in a DR situation everything sharing an OBR6 pool is perfectly smart and reasonable.
-
@hubtechagain said:
WTF is starwind?! ha
This is who we are
https://www.starwindsoftware.com/starwind-virtual-san-free
We give away free version to use on a bare metal servers (so you take a pair of them and turn into HA NFS or SMB3 NAS). This one has no restrictions at all (capacity is unlimited, production use is OK and you can be anybody to get it).
HA iSCSI and hyper converged versions are available to different set of people like MVPs, SpiceHeads, VCPs, some restricted ones to MCTs & bloggers. Technically we can bring same program to MangoLassi community as well. I just need some sort of a low watermark (points, rank or whatever) to make the program look a bit of private so my VP of Sales would not burn me with a blow torch
Cheers and let me know if you'd have any questions
-
@hubtechagain said:
using the hyper v replication, how do the VMs behave? they're offline, replicating, then boom, tree crashes through building and smokes our server rack. i just remote into the DR server and spin em up?
Yup. You can even automate this process with PowerShell (not recommended as there's no protection against brain split) and use Azure orchestration site (this one is OK). See:
(Keith is a VERY smart and exceptionally pleasant guy BTW)
(Robert is MSFT serial Cluster MVP and I'm happy I know him in RL)
Good luck
