@scottalanmiller said in Veeam with NetApp?:

But there are way faster options for less money, making it a moot benefit.

Do you mean as a total solution or a faster NAS?

@scottalanmiller said in Veeam with NetApp?:

NetApp makes high cost, low performance NAS devices on double parity RAID 4 (basically RAID 6 but using older RAID 4 tech instead of RAID 5 based.) They could be talking about any NetApp product as that's basically what they all do. You just pick one based on size.

I can't think of any situation where I'd want to do this, though. NetApp has strengths, but extremely few, and this isn't one of them. The reason that people chose NetApp for this is because they want RAID 6 with spinning disks but the write penalty is too high and RAID-DP while not quite as safe and not as fast for restores, is slightly faster for writes. But there are way faster options for less money, making it a moot benefit.

OK, thanks. So in it's basic form they are talking about VM snapshot backups to a NAS?

I was reading about a company where they mentioned that they are setting up one or a few VMware hosts on-prem and using Veeam with NetApp as a backup solution. And that's all the info I have.

I was wondering what kind of backup solution / products they are likely to be talking about?

200 Mbps is not bad. I'd be happy with that.
You'll probably find out that the speed will vary a lot from day to day and hour to hour.

Best performance will be when you use the same ISP and are physically close.
It's also possible that the ISP will throttle your connection automatically depending on what port/protocol it sees. If so then running on another port sometime gets around that.

Also if we look at density from 5 years ago, drives have gone from a max capacity of 6TB to 16TB.

@scottalanmiller said in Spinning rust, how long do you keep it spinning?:

@Pete-S said in Spinning rust, how long do you keep it spinning?:

Economical lifespan. Either run it until it fails or plan to replace it when it has done its duty and take advantage of 5 years of technology advancement.

Is that really an economical lifespan, though? We normally think of drive economy as being best when we keep them far longer than that. And there is almost no tech advancement in just five years in drives. For most companies, this would roughly double the cost of drives as they normally get ten years from them, rather than five.

Well, 5 years is what you can be sure you will get with a 5 year warranty. It wouldn't make sense to use anything longer when doing the investment.

Look at Backblaze for instance, the oldest models they have in use (Q2 2019) are 51 months old. That's a little more than 4 years. And they are very much into keeping costs as low as possible.

@scottalanmiller said in Spinning rust, how long do you keep it spinning?:

@Pete-S said in Spinning rust, how long do you keep it spinning?:

All enterprise drives have 5 year warranty. 24/7 that means 43800 hours. So my thinking is that when the drives hits that many power-on hours, they should be flagged for retirement.

What's the connection?

Economical lifespan. Either run it until it fails or plan to replace it when it has done its duty and take advantage of 5 years of technology advancement.

The recycling market will decide what economical value your 5-year old drives have.

For our own use, the days of hardware raid are over. Only software raid. And we don't put the OS of the host on it. So that makes the array easy to move and not at all coupled to the hardware of the box.

All enterprise drives have 5 year warranty. 24/7 that means 43800 hours. So my thinking is that when the drives hits that many power-on hours, they should be flagged for retirement. Power-on hours is also easy to manage since it's in the smart data.

Most drives in the same array will be of the same age. So often that means moving the data to a new array or maybe store it some other way. 5 years is also long enough that technology has progressed.

@mlnews said in Miscellaneous Tech News:

Debian 10: Playing catch-up with the rest of the Linux world (that’s a good thing)

If you skipped the last release, Debian 10 (Buster) should encourage an update into 2019.
The Debian project, the upstream mother of countless Linux distributions, has released Debian 10, also known as "Buster." And yes, that's a reference to the character from Toy Story. All Debian releases are named after Toy Story characters. Bla bla bla...

WTF is news about that? Debian 10 been out since July.

Next release will be always be in about two years or so. It's released when it's ready and that's how it has been for almost twenty years now.

If you want to post real news then it's that Debian 10.1 was released a couple of days ago.
https://www.debian.org/News/2019/20190907
You can still use old 10.0 iso install media and when you update packages you will get the same as 10.1.

This is the latest netinstall image Debian 10.1 for amd64:
https://cdimage.debian.org/debian-cd/current/amd64/iso-cd/debian-10.1.0-amd64-netinst.iso

@CCWTech said in Server with multiple backplane / Drive Configuration:

@Pete-S said in Server with multiple backplane / Drive Configuration:

@CCWTech said in Server with multiple backplane / Drive Configuration:

@Pete-S said in Server with multiple backplane / Drive Configuration:

What server/model is it?

HP ML350 Gen 10

https://techlibrary.hpe.com/docs/iss/ML350_Gen10/setup_install/

OK, looks like there are many storage options on that machine and many different raid controller options as well.

But, there are no SAS expanders on the backplanes. Backplanes are straight "dumb" wiring so to speak. HPE has an additional SAS expander card that have to be installed if you're using that.

So if you have 16 drive bays I assume 8x2.5" drive bays per "box". You then probably have it paired with the P816i-a controller in which case you have each drive bay directly wired to one SAS port from the controller.

So it makes no performance difference how you install the drives. I would however put the SSDs in one box (one backplane) and the HDDs in the other - just to make it less confusing and less of chance to pull the wrong drive when something fails.

I don't understand why you would use 2.5" HDDs at all though but that is another matter. The server can also be configured with one 8x2.5" and one 4x3.5" bay which would be my preference.

Yes, it is the HP SAS expander card I'm using.

Since the backplanes in that server doesn't do anything, it makes no difference in what bays you put the drives. Not from a performance perspective and not from a redundancy perspective.

From the SAS expander to the backplane to the drive is electrically the same as just having a straight cable. So there is nothing that can fail - unless you physically abuse it.

The backplane does have some electronic components on it though but that is for driving the LEDs in the drive bays.