@scottalanmiller said:

I think that the V100 is the big "find" of the set because it is small and easy to ship, quiet to run in a house, can be racked or easily wall mounted, is light and easy to move around yet provides a very cool non-PC server platform. Runs FreeBSD well, has two drives for RAID 1 but is Sparc based and is perfect for forcing you to gain perspective on using non-Intel and non-AMD architecture systems.

Absolutely love Sun hardware! Still have Ultra10 and Ultra40 in a running conditions somewhere in our basement.

P.S. It's a pity Sun is out. Pretty much like DEC.

@Dashrender said:

@KOOLER said:

@Dashrender said:

I wasn't aware that PCIe SSD were as reliable as RAID based SAS SSD until I ran across a thread on SW this morning.

These devices are (or at least were) normally more than 100% more than the cost of similar sized SAS SSD and the accompanying RAID controller.

Now obviously we IT folks in general have found that single RAID controllers are reliable enough to not warrant having a backup within the same chassis, we find that it's just as likely to have a whole die as to have the RAID controller die, so instead of backing up the RAID controller we backup the whole system to cover those situations.

I've done a little reading now, it seems that the resiliencies of PCIe SSDs are approximately equivalent to a RAIDed setup, and the controller is at least equivalent to normal RAID controllers.

As the costs come down, these seem like the clear winner.

What is your experience?

We recommend using PCIe and newer NVMes for cache. You can go all-flash of course but that's more expensive than desired typically. Back to SAS SSDs they are overpriced IMHO compared to their SATA siblings.

Do you recommend SATA SSDs over SAS SSDs because the price point vs performance/failure rate rarely warrants it?

Scott had replied below. Key point with commodity hardware is - economy of scale kicks in, price goes down dramatically so typically more expensive stuff has not that great $/TB or $/IOPS rate compared to COTS one.

@Dashrender said:

I wasn't aware that PCIe SSD were as reliable as RAID based SAS SSD until I ran across a thread on SW this morning.

These devices are (or at least were) normally more than 100% more than the cost of similar sized SAS SSD and the accompanying RAID controller.

Now obviously we IT folks in general have found that single RAID controllers are reliable enough to not warrant having a backup within the same chassis, we find that it's just as likely to have a whole die as to have the RAID controller die, so instead of backing up the RAID controller we backup the whole system to cover those situations.

I've done a little reading now, it seems that the resiliencies of PCIe SSDs are approximately equivalent to a RAIDed setup, and the controller is at least equivalent to normal RAID controllers.

As the costs come down, these seem like the clear winner.

What is your experience?

We recommend using PCIe and newer NVMes for cache. You can go all-flash of course but that's more expensive than desired typically. Back to SAS SSDs they are overpriced IMHO compared to their SATA siblings.

@Dashrender said:

Of course if your talking about doing starwind between the business site and the DR location, what kind of bandwidth do you need?

I know Hub has 10 mil ping times.

Don't do it synchronous way. Not even needed to beg for troubles. Bandwidth is not free and unless these are two buildings on same campus I see it as an overkill: I'm personally not a big fan to replicate whole sites or virtual LUNs. Smaller objects like running VMs (Hyper-V Replica or Veeam Replication) are better candidates for that.

@Dashrender said:

StarWind isn't possible in this setup because of the lack of enough local storage on the two servers in the same location.

You can go virtual on top of an exiting hypervisor nodes. There's a way to obtain free license for hyper converged setup if you plan to support and maintain everything on your own. FYI.

@Dashrender said:

@KOOLER said:

@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

That is pretty cool. Most of us here are Spiceheads as well, so we're probably covered, though getting ML on the list would be awesome!!!

It's still a smaller community here as with a set of drawbacks it definitely has own benefits: much easier to have "special" handling

P.S. You know bigger you become more bureaucratic processes start to happen to slow the things down and complicate everything...

@mlnews said:

Pinging @KOOLER

Replied! Thank you for brining me in

@Dashrender said:

Assuming both nodes in a two node setup are the same. Is it better to split the VM load between the servers, meaning Starwind has to replicate in both directions, or is it better to run them all from a single host replicating in one direction?

You create two virtual LUNs and you split your VMs between two hosts having own virtual LUN replicated to "shadow" one on a partner. Within this scenario there's minimal "fight" for any of LUNs ownership so performance is better.

@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:

http://blogs.technet.com/b/keithmayer/archive/2012/10/05/automate-disaster-recovery-plan-with-windows-server-2012-hyper-v-replica-and-powershell-3-0.aspx

(Keith is a VERY smart and exceptionally pleasant guy BTW)

https://robertsmit.wordpress.com/2015/02/18/azure-site-recovery-provider-for-hyper-v-how-to-replicate-hyper-v-to-azure-recovery-cloud-hyper-v-winserv/

(Robert is MSFT serial Cluster MVP and I'm happy I know him in RL)

Good luck

@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

FreeBSD R10

Can't make myself love Linux. Trying once in a while since 1993 or 1995 and no luck so far

@ajstringham GOTO is for script kiddies! Real man with a big brass <...> use "jmp" or "B".

@Addie Reputation should either go both directions OR keep "good" and "bad" karma counters. IMHO