Re: [PLUG] Is it possible to move data from a SAS drive to the newly installed SSD?

[H Mottaleb via plug <plug@lists.phillylinux.org>]

My goal is to run an Eth2 node and have it running 24/7. As far as the downtime, within a day should be ok. RAID would be good but it would cost quiet a bit to get several ssds. I have a mini computer with with an i7 cpu and a ssd and I can get another one to backup the data along with the OS. 

Is there a way to copy everything from the server on to the ssd so it continues on the new system? Maybe mirroring the hd? And, would I be able to backup everything once it’s all synced and fire it up if there’s an issue with the current setup?

Sent from Yahoo Mail for iPad

On Sunday, January 17, 2021, 9:07 PM, Rich Freeman <r-plug@thefreemanclan.net> wrote:

On Sun, Jan 17, 2021 at 8:15 PM H Mottaleb <h_mottaleb@yahoo.com> wrote:
>
> So, i don’t have to have the raid setup if I continuously back it up? How about moving the data to another system with a ssd?

So, other than running geth, what exactly is your goal here?  I don't
want to tell you to build a system out of rubber bands and string and
then find out that you are moving $10M/day through it as that is a
problem waiting to happen.

If this is just a fun little project where the ultimate culmination
will be using some ETH to order $20 worth of stuff off of whatever
Amazon-like-company accepts ETH, then there are lots of options, and
really it is no big deal if something breaks.

If this is something serious then there is a way to go about it seriously.

What is the cost of downtime?  If you lost your host, and the solution
was to restore a backup which rolled back your copy of the blockchain
by a month, and your node is unable to service transactions for a day
or two, is that a big deal?

RAID is fundamentally a solution to downtime due to lost disks.  All
disks fail.  RAID greatly reduces the risk of a disk failure causing
downtime, at the cost of added complexity and extra disks.  If you
can't tolerate downtime, then you need RAID or some equivalent
solution to storage hardware failure.  Of course, if you can't handle
downtime you also have to consider that every component in that server
is subject to failure.  Since ETH is already distributed by its nature
you might actually get further with two really cheap independent hosts
than a really expensive single-host enterprise system.  The blockchain
already replicates itself - the only thing you'd need to worry about
are any wallet files.  The only time the wallet changes is when you
create a new account - transferring money is a read-only operation as
far as the wallet is concerned ( the wallet is just a fancy name for a
private key ).  So if you aren't creating new accounts often then you
can just backup your wallet every time you do so before putting any
money in the new account, and then everything on your server is
disposable with the only cost being downtime to replicate the
blockchain again.

If your goal is to learn more about enterprise hardware/etc we can do
that too - though obviously that tends to cost more money to get the
same job done.  Keep in mind though that software-level redundancy
with cheap hardware is a legit strategy in the real world - really big
companies do just that.  It just requires the right software.  It just
so happens that geth itself is one of those services well-suited to
this.  And the questions I asked above are one of the first steps you
want to answer when designing anything for the enterprise.  It isn't
enough to buy expensive server hardware - you need to understand your
uptime, recovery time, recovery point, etc objectives before you just
assume that even expensive hardware will deliver.

--
Rich

___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug