John
Member
Registered: 30th Jun 03
User status: Offline
|
Just had a quick read over this what pow has said is correct, if you've just got the 2 shares open, it'll go via your PC.
However, is that 1.5TB correct? It would take a lot longer than 3 days to move that via your PC.
|
willay
Moderator Organiser: South East, National Events Premium Member
Registered: 10th Nov 02
Location: Roydon, Essex
User status: Offline
|
without knowing specifics about the connection between the UK and US office its hard to judge.
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
quote: Originally posted by John
However, is that 1.5TB correct? It would take a lot longer than 3 days to move that via your PC.
Yeah sure is? How do you know without knowing what the connection they have is? Curious.
Unless you mean because it was to the US? The US bit might not necessarily be the case, I just used that for effect to make it simpler to explain. The NAS is definitely at another site, but it could just be in the UK.
[Edited on 21-02-2014 by Brett]
|
John
Member
Registered: 30th Jun 03
User status: Offline
|
That's just on the assumption that everyone isn't using proper connections. If they are, it's possible.
We've got some sites with a PWAN between them which acts like a switched network, could move 1.5TB over that on a 10meg connection over 3 days.
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
This is a massive aerospace company worth billions where all data is spread across multiple sites, so I'm guessing they spent the money on a decent connection? Just with how it's setup how I described, when it comes to TB's at a time, it can take a fair while.
[Edited on 21-02-2014 by Brett]
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
You'd be surprised Brett
I'm surprised the 1.5TB manages to copy without dropping it at some point
|
willay
Moderator Organiser: South East, National Events Premium Member
Registered: 10th Nov 02
Location: Roydon, Essex
User status: Offline
|
it will drop but SMB/tcp will recover
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
That's the other thing, simply doing a simple drag and drop in explorer, it has and there's no way to resume so they just start again? Though that has happened in the past supposedly, it has been a seemingly stable transfer since I've seen it done. Or they get this problem where the amount of sub folders, filenames, etc, when copied over then exceeds the maximum file length in windows and causes a problem, 256characters is it or somethin like that? Guessing the Archive folder is also within sub folder upon sub folder
I'm pretty much launching a campaign to sort this now as it's complete bullshit and could be setup more efficiently I'm sure They get to a point where the drive is filling up so stuff HAS to be archived away, else if not all work stops basically. Stupid.
[Edited on 21-02-2014 by Brett]
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
Fucking drives filling up in this day and age. I just hit the buffers on a 1TB VHD, so I shut the VM down, took a backup of the VHD, expanded the disk, fired the VM back up and extended the volume. Without the backup it took oo... 3 minutes to complete?
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
I thought that with regards to space, but this is kind of more data than I've ever seen. They're knocking out 100's of GB's a day.
[Edited on 21-02-2014 by Brett]
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
Then surely they should have some sort of SAN to deal with it
|
Rob_Quads
Member
Registered: 29th Mar 01
Location: southampton
User status: Offline
|
quote: Originally posted by pow
Fucking drives filling up in this day and age. I just hit the buffers on a 1TB VHD, so I shut the VM down, took a backup of the VHD, expanded the disk, fired the VM back up and extended the volume. Without the backup it took oo... 3 minutes to complete?
This winds me up rotten at our place. Because anything in the server room is 'server grade' I just wanted some extra space but it costs a stupid amount. 600GB disk for $1000. Ask for an SSD and its $4000 for 100MB.
Most of the time we don't care if we loose a disk (and from out experience I would say we loose server grade discs just as often as ones on desktops / home
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
Eh? I know I'm not exactly high end here but I'm using 2TB WD Blacks in our server RAID arrays. SAS storage is a joke price wise though
|
Dom
Member
Registered: 13th Sep 03
User status: Offline
|
Backing up what others have said - it's unlikely your IT department would grant you direct RDS/VNC/'Remote' access to the storage server, rather you want RDS/VNC/'Remote' access to (any) a machine that resides on the same local network (State side) that has those specific network shares attached preventing a 'transatlantic loop', slowing the transfer speed, when transferring data between the (State side) shares this side of the pond.
Essentially.....
[Edited on 21-02-2014 by Dom]
|
Rob_Quads
Member
Registered: 29th Mar 01
Location: southampton
User status: Offline
|
Yeah all our stuff is SAS stuff.
Were just pricing up some SSDs for one of our performance boxes and we are looking at around $12,000 to get 3 bloody discs to put into a RAID0 array. These are the sort of discs we are looking at - http://www-03.ibm.com/systems/x/options/storage/solidstate/enterprise.html
|
John
Member
Registered: 30th Jun 03
User status: Offline
|
At least SAS disks in any hald decent servers, normal SATA drives don't cut it.
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
quote: Originally posted by Dom
http://i.imgur.com/xb0gX65.png
Nice, I'll use something like that House brick might've been more appropriate for 'brett' tho, especially in this instance Is transatlantic loop a proper term? Like that, will make me sound like I know what I'm on about lol
Big thanks to everyone who's responded, feel confident about what it is I'm after now
[Edited on 21-02-2014 by Brett]
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
Dom that picture is amazing
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
One more thing... I guess once on the same LAN, the limitations then will only be local networks speed, etc?
I guess as a test case I could make a share on my machine and the machine next to me, then transfer over data from one another? Any way of recording it so there's figures I can use to show the difference in transfer speeds and make a 'vs' kind of chart/graph?
|
Rob_Quads
Member
Registered: 29th Mar 01
Location: southampton
User status: Offline
|
Has to be said that is an awesome picture. Provided much humor to much of my team
|
John
Member
Registered: 30th Jun 03
User status: Offline
|
http://iperf.sourceforge.net/
|
Dom
Member
Registered: 13th Sep 03
User status: Offline
|
quote: Originally posted by Brett
Is transatlantic loop a proper term?
If by proper you mean bullshit then yeah, it's a proper term Can always throw in 'induced latency' for extra finesse
quote: Originally posted by Brett
One more thing... I guess once on the same LAN, the limitations then will only be local networks speed, etc?
In theory yes. Although the remote system wants to be as close to, in terms of network routing, the storage server as possible if planning to use Samba (SMB) network shares.
Saying that it might be worth asking if they're able to auto archive specific directories when they reach a certain size threshold - would save a lot of peeing around from your end!
|
Brett
Premium Member
Registered: 16th Dec 02
Location: Manchester
User status: Offline
|
quote: Originally posted by John
http://iperf.sourceforge.net/
Sounds like it might've done the job, but looks like it's been removed
quote: Originally posted by Dom
If by proper you mean bullshit then yeah, it's a proper term Can always throw in 'induced latency' for extra finesse
The more bullshit terms, the better tbh I'm trying to convince management who don't have a clue that I know better than IT here
quote: Originally posted by Dom
Saying that it might be worth asking if they're able to auto archive specific directories when they reach a certain size threshold - would save a lot of peeing around from your end!
In an ideal situation, yes. Unfortunately it's done this way because the archiving doesn't necessarily have any relevance to how old it is, more so about whether it's still in use. Could very well end up where newer stuff is archived over stuff that's months old. I will also be investigating if there's a more automated means to knowing when things are done with tho. Even something as simple as sticking an X on the end of the folder, but then it has the problem of relying on people to do that, which I know they won't. Could be another way tho, but one step at a time
|
VrsTurbo
Premium Member
Registered: 8th Jun 10
User status: Offline
|
quote: Originally posted by pow
Eh? I know I'm not exactly high end here but I'm using 2TB WD Blacks in our server RAID arrays. SAS storage is a joke price wise though
i hope your only using 2tb drive for archives or large data stores.
|
pow
Premium Member
Registered: 11th Sep 06
Location: Hazlemere, Buckinghamshire
User status: Offline
|
Yeah just data, my SQL databases are on a SAS array
|