Very Slow Transfer Rate Solved


Author
Message
GDK
GDK
New Member
New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)
Group: Forum Members
Posts: 4, Visits: 12
I've been experiencing very slow transfer rates during File and Folder backups, but have finally figured out the problem. This is a quad core 4.1 GHz WIN7 desktop with an SSD and several HDDs. All internal eSATA drives. During the backup process from either the SSD (Drive CSmile to the backup HDD, or from the (Drive DSmile HDD to the backup HDD, I would see the transfer rate start out around 500 Mb/s or so, but then over the course of a few minutes watch it decay to something like 100 Mb/s, eventually stabilizing at about 12 Mb/s! That's unusable.

I turned off the firewall, killed the AV, disconnected the ethernet cable to the router, and stopped every process not relevant to the backup, all with no discernable effect. Finally, on a hunch, I disabled my Logitech wireless mouse by pulling the dongle out of the USB port. Within seconds the transfer rate went to 1.0 Gb/s and stayed at or near that rate! 

Don't know if this will help anyone else, but it might be worth a try.


jphughan
jphughan
Macrium Evangelist
Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)
Group: Forum Members
Posts: 4K, Visits: 29K
USB 3.0 (from your hard drive) and 2.4 GHz wireless signals (which Logitech uses) are known to interfere with each other, which is why some wireless routers with USB 3 ports place them at inconvenient locations such as on the front of the router to move them as far away from the antennas at the rear as possible (to the confusion and frustration of many who aren't aware of this, including tech reviewers....), so that actually doesn't surprise me.  If you've got a USB port on the opposite side of your laptop/tower, they should be able to coexist peacefully. Smile

Froggie
Froggie
Master
Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)
Group: Forum Members
Posts: 937, Visits: 7.9K
He stated all his drives were INTERNAL (SATA)... they should not be affected in any way by "wireless" mouse signals.  This sounds more like a driver interference/priority issue at a low System level.
Edited 4 May 2017 8:25 PM by Froggie
Richard V.
Richard V.
Most Valuable Professional
Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)Most Valuable Professional (4.1K reputation)
Group: Forum Members
Posts: 2K, Visits: 8K
@GDK -- "All internal eSATA drives."?  That's a bit confusing.  Unless you say otherwise, I would assume that you meant that they're all on internal SATA (not eSATA) ports.

What USB port (USB2 or USB3) was/is your Logitech dongle connected on?  AFAIK Logitech's current version's own capabilities are USB2 only, although it should work without problems on a USB3 port that is fully backward compatible.  Despite manufacturers' optimistic claims, not all of them are, especially front panel add-ons.  In any case, it would be worth testing results when connected on a different USB port if one is available.  I've seen no impacts on Reflect backups similar to what you've reported, but I have seen some mouse and keyboard hesitations with various "Unifying" receiver connections on my own system.

Regards, Richard V. ("Arvy")
https://forum.macrium.com/uploads/images/afc5d4fe-5d25-4e25-be94-185e.png

Edited 4 May 2017 8:46 PM by Arvy
jphughan
jphughan
Macrium Evangelist
Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)
Group: Forum Members
Posts: 4K, Visits: 29K
Sorry, missed the note about all internal drives -- but yes, "internal eSATA" is a bit confusing, although even if they're external eSATA drives I'm unaware of interference issues between 2.4 GHz and eSATA.  Then again, I haven't looked into it since eSATA never very widely adopted (lack of a standard means of providing power on that cable was a big annoyance) and now we have USB 3.0 and Thunderbolt.

Froggie
Froggie
Master
Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)
Group: Forum Members
Posts: 937, Visits: 7.9K
Most eSATA connections were passive (no power issues) and only differed from SATA by connector and cable length.  I had lots of Systems that only had li'l backpanel connectors (eSATA FEMALE) that connected passively to the internal SATA connectors... no power issues.
jphughan
jphughan
Macrium Evangelist
Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)
Group: Forum Members
Posts: 4K, Visits: 29K
Froggie - 4 May 2017 9:39 PM
Most eSATA connections were passive (no power issues) and only differed from SATA by connector and cable length.  I had lots of Systems that only had li'l backpanel connectors (eSATA FEMALE) that connected passively to the internal SATA connectors... no power issues.

I wasn't talking about power issues.  My point was that the standard eSATA connector didn't provide power AT ALL because it only included pins from the internal SATA data connector, not the additional power connector that internal SATA disks require -- and that made eSATA unsuitable for form factors like 2.5" external drives where the expectation is that they'll be bus-powered, as they are when they use USB.  For 3.5" drives that require separate power anyway, yes eSATA was a great alternative to USB 2.0, but by the time eSATA came around, 2.5" drives offered enough capacity for "most people" and they liked the portability and single cable convenience.  A few vendors did indeed include eSATAp connectors that offered power (via USB power pins in eSATA+USB combo ports), but that was never widely adopted because it came too late.  The only eSATAp device I remember using was a Dell external optical drive that was an accessory specifically for a Dell laptop my wife bought that didn't have one built in.

Edited 4 May 2017 10:16 PM by jphughan
GDK
GDK
New Member
New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)New Member (6 reputation)
Group: Forum Members
Posts: 4, Visits: 12
Sorry, I meant SATA drives. The mouse dongle is actually plugged into a externally powered USB hub which is connected to a USB2 port. The motherboard (Gigabyte X58A-UD3R) is a few years old. Maybe some driver issues?
Although I've got the internal drives working well, I can't get my external 500GB WD Passport (USB3) to go better than about 35MB/s. But, upon completion Macrium shows, I/O Performance:Read 1.8 Gb/s - Write 739.6 Mb/s seems pretty fast for what I'm actually seeing.

jphughan
jphughan
Macrium Evangelist
Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)Macrium Evangelist (5.9K reputation)
Group: Forum Members
Posts: 4K, Visits: 29K
Well first, be aware that to convert Mb to MB you have to divide by 8, so 739 Mb/s write is 92 MB/s.  Even taking that into account though, I too have seen implausibly high speeds reported by Reflect, so I'm wondering if that rate is measured by the amount of source data being read/written rather than the amount of data being read/written at the destination -- which can be a very different figure if you're using compression and achieving a decent ratio.  Even then though, unless you're copying a lot of tiny files, 35 MB/s write to a USB 3.0 disk is pretty bad, even a smaller one like 500GB that would likely have lower platter data density.  Is that the best you see even while copying a single large file, such as a Windows Install ISO?  If you're on Windows 7, do you have the latest Intel USB 3.0/xHCI drivers?  I know they've been revised quite a bit over time.  Windows 7 also doesn't perform as well as newer versions because it doesn't support UASP, but I'd still expect better than 35 MB/s.  And this is a much longer shot, but did you by any chance originally format that disk on a Windows XP machine?  If so AND the disk uses 4K sectors (also rather unlikely for a 500GB disk), Windows XP sets disks up such that the logical and physical sectors are never aligned, which dramatically reduces I/O performance.  The easiest fix is to reformat the disk on Windows Vista or newer, or if that's not an option there are utilities to "re-align" the data on a disk.

Otherwise, are you sure you're using a 3.0 cable and USB 3.0 port?  Have you tried another cable and port even if you are?  If you're using a port that internally connects to a pin header block on the motherboard, are you sure it's connected to a USB 3.0 block rather than 2.0?

Edited 4 May 2017 11:51 PM by jphughan
Froggie
Froggie
Master
Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)Master (1.6K reputation)
Group: Forum Members
Posts: 937, Visits: 7.9K
GDK, on that USB3 issue, you might wanna shop around on USB3 drivers for that System/hardware... it looks like its using a USB2 driver.

The Reflect #s are a little bit bogus... try not to take them too seriously unless the Devs tell you what they really mean.  To me, it looks as though they may represent the DATA written BEFORE compression is applied, which of course would be about DOUBLE the actual DATA rate.

Only they know the real answer...
GO

Merge Selected

Merge into selected topic...



Merge into merge target...



Merge into a specific topic ID...




Similar Topics

Reading This Topic

Login

Explore
Messages
Mentions
Search