Any way to speed up network backup

Possibly, but it may mean some hardware changes.

Here's my experience you can use as a reference. I backup 3 computers running Win7 32 bit to a Western Digital My Book World Edition 1 TB NAS. I'm connected via a Linksys WRT610N router (wireless n & gigabit). I have gigabit cards in the 2 desktops and a wireless n card in the laptop. I use TI Home 2010 and only do full system/disc images to the NAS. I have True Image set to use normal compression and normal backup priority.

The two desktops transfer data at about 15 MBps (that's bytes, not bits) to the NAS during a backup. The last system image I did was 88,288 MB (86 GB) and it took 96 minutes to run, so 15.3 MBps. The choke point in the system is the NAS drive itself. According to WD information the NAS is performing as expected with such a large file size. I get up to 25 MBps with small files. So with my network and your file size of 329 GB (336,896 MB) a backup would take 374 minutes, or almost 6-1/2 hours.

A few more numbers for reference. Prior to upgrading to gigabit, I was using the 100 mbps network and achieved speeds of about 9 MBps. I got a 50% increase going to gigabit. If I had a more robust NAS I could get higher rates. If your network is running at about this speed of 9 MBps the 329 GB backup should take 10.5 hours.

When I copy files from one desktop to the other over the gigabit network I get about 30 MBps transfer rates. From the wireless N laptop I get about 10 MBps. I don't ever get higher than 30 MBps since my desktops are 4 and 5 years old and there are other system limitations within them that keep me to that rate.

When you say that you're 50% done, does that include verification? If so the backup itself could be just about done. Also, I've found the counter in the system tray for the new version of TI 2010 is unreliable. It always tells me its only 60% done or so just before it finishes.

Hope that helps.

The only way to accurately time TI is with a clock or after the task is done look at the times in the log. Ignore the progress bars.

The posted numbers for networking hardware are the absolute maximum signaling rates possible between the NICs and do not account for any overhead at all. In other words, you won't get it in real life. Just for a theoretical comparison, a USB drive is 4.8X faster than your 100Mbps network.

On PCs with some type of P4 processor, the rule of thumb for an image to an internal drive was to allow about 1GB of compressed archive per minute. If I assume compression on your partition to be 0.8 then your compressed archive would be about 263GB and at a minute/GB it would take 4.3hrs - and that's to an internal HD.

Hmm, well I can definitely copy data to and from the NAS quicker (its a Netgear ReadyNAS in Raid 5). When I watch it copy data to the PC normally, the activity light blinks constantly, whereas when I was imaging the drive it just flickered occasionally. Would turning compression down speed things up perhaps? The entire operation took around 17 hours in the end.

Turning off compression will reduce the load on the CPU but will cause more data to be transferred over the network since the archive will be larger.

Try it, it may point out where the bottleneck really is with TI on your system; processing or transferring.

Try adjusting the priority in the backup wizard options to high if it is set to normal or low and see if that helps.

This thread is about the closest discussion to what I'm doing and in fact, Peregrine's setup is very close to what I've been working with at home. I also use a WD MyBookWorld White 1TB NAS which has a 1Gb port. This drive and my main computer are also on 1Gb connections through a Netgear 1Gb switch hooked to a Netgear WNDR3700 1Gb router. I also have other various wireless and wired components, but those aren't part of the issue here.

I've been trying to better understand data transfer rates, how to measure/monitor them, and how to improve them, both across my network and just from my main computer to other directly connected devices, such as a new 2TB Fantom eSATA drive we've just added. The Fantom runs a Samsung HD204UI drive (7200RPM) and is connected directly to a PCIe eSATA card in my main computer and is my backup destination drive. The source drive in my computer is an internal SATA 1TB WD10EADS-65L5B1 (7200RPM). All in all, I think this should be a fairly fast environment for Acronis to do it's job and I believe my numbers are pretty good. I'll get to that in a moment.

I have a couple different concerns to address in this post, but my main overarching concern is that, even with all of the hardware improvements I've made...upgrading my router and network to 1Gb capability, upgrading my backup drives, upgrading my main computer....I'm still not sure that I'm not experiencing data transfer bottlenecks in my system or network and I'm not sure that Acronis is performing correctly or efficiently.

I use a couple of monitoring tools, one reports network transfer speeds and the other I just recently found which shows me the read/write speeds of specific disks in my system. The network monitor is great for clocking transfers to/from the WD MyBook drive, which is the backup destination for all other computers on my network. The program I use is called the "Net Meter". It's from http://www.hootech.com and it gives a nice graph for throughput on each network adapter you have installed....very handy. The other program I found for monitoring individual disk read/write speeds is called "HD Tune Pro v4.6" and should be easy to find. I would really appreciate hearing about any other software tools and/or techniques that you folks are using which can help us all better understand what is going on with network/device data transfer in real time, so please share if you can.

Ok, my primary issue stems from a full backup I did just this evening of my main computer's disk. According to the Win7 ledger in MyComputer, I have used 223GB from my available 931GB on the WD internal SATA. I setup a new incremental backup schedule today and started fresh after installing the new eSATA card and Fantom 2TB external drive combo. Backups had been previously going across the network to the WD MyBook drive, but I wanted to see if there was any speed improvement going from the 1Gb network connection to the direct eSATA setup. I also downloaded and installed the latest build 6597 of ATIH2011.

The results were muddled, unfortunately, because I had some "issues" with the backup. First, I had a sector read error, which has NEVER happened before with this WD10EADS source drive. I can't help but suspect that the new ATIH build 6597 had something to do with that. Here is the error message text:

13 Error 12/10/2010 4:48:59 PM Failed to read data from the disk.
Failed to read from sector '103,730,392' of hard disk '2'.
Failed to read the snapshot. (0x10C45A)
CRC error (0x100155): ⎂Retry/⎂Ignore/Ignore ⎂All/⎂Cancel

I canceled and retried the backup several times, but kept getting the same error...same sector. I also ran Chkdsk and no error was found. After that, I decided to "ignore" the error and let the backup proceed. There were no further errors after that and the validation completed ok too. However, the final archive file ended up being over 600GB...and my original data was only 223GB!! This was a new, full backup, with normal compression, unencrypted, sector-by-sector without backing up the unallocated space. Ok, that's ONE mystery...I've never had TI come up with an archive file LARGER than the original source data for the many years I've been using it. What would/could cause that? I may post this issue separately due to it not quite being relevant to the particular topic, but I wanted to mention it here because it was part of this speed monitoring process.

Ok, now for the next part of this which is more relevant to this thread. I was very intrigued by the monitoring results of the eSATA drive during the read/write functions of backup and validation. The HD Tune Pro trial software I used has a monitoring function that graphs read/write speeds in real time and it also has a screenshot function. I took several shots during the entire backup/validation process which took 5 hours and 50 minutes. I guess that isn't too bad for a 600GB backup/validation process is it?

Anyway, what I noticed from the HD Tune graphs was that the ATIH backup process did not occur in one continuous feed of data to the destination drive. There was a LOT of dead space in between drive writes for several minutes....then other times where the write WAS continuous for several minutes. When ATIH was actively writing data to the eSATA disk, it was happening very fast at a sustained rate of between 50MB/s-70MB/s. The writes did seem to be in some sort of repeating pattern, but I would like someone from Acronis to perhaps comment on these things. Particularly, the fact that ATIH2011 does not seem to be able to process backup data fast enough to keep it flowing continuously to the destination drive, at least in my case.

Also, you will see in my pictures attached that when the validation process begins, the READ process IS completely continuous without any "dead space" at all. Further more, the speed was phenomenal from the Fantom eSATA drive, easily maintaining sustained speeds of over 100MB/s with a maximum of about 125MB/s for almost 90 minutes of validation. If the backup WRITE process could be made continuous, then the entire 600GB backup AND validation would have been MAYBE about 4 hours. Again, I don't think 6 hours is too shabby, but I'm just wondering if the non-continuous write patterns during backup represent normal patterns/activity for ATIH or if there is perhaps something else going on that should be addressed.

See pics attached...the orange sections are backup WRITES...and the blue is the validation READS. Thanks!

I just wanted to add that the important thing for me with all of this is reducing backup times as much as I can. I intend to do some additional testing to get as close as I can to an apples/apples comparison of the backup times between the 1GB ethernet connected WD MyBookWorld drive versus the Fantom eSATA connected drive as backup targets. At some point, I will likely add USB 3.0 capability and see how that also figures into things.

From the Acronis USER community, I hope to see if anyone is doing anything along these lines and if so, perhaps we can compare notes. From Acronis, I would like to know if you think the intermittent WRITE patterns displayed in my screenshots from the previous post represent "normal" functionality for ATIH2011. If not, then I need to investigate what we can do to resolve it. If those patterns ARE normal, then my question to Acronis is....can this be improved to speed up the backup write time and make it more like the validation read time. It appears that validation times are pretty optimal with no "gaps" in the process.

Thanks!

This was a new, full backup, with normal compression, unencrypted, *sector-by-sector* without backing up the unallocated space. Ok, that's ONE mystery...I've never had TI come up with an archive file LARGER than the original source data

Yes, sector-by-sector backups are large, 60% of total disk size is something to expect. And it probably became such after the "ignere"-d sector read error.

Thanks for your response, but I do not understand your logic here. First of all, I only ran the backup process to it's full completion AFTER a CHKDSK found NO ERROR and STILL got the sector read error from Acronis, so what am I to do? I'm now running ANOTHER CHKDSK function this morning just to double check.

Secondly, what is the purpose of the tick box that allows you to include/not include "unallocated space"?? Perhaps I'm mistaken, but that at least implies to me that a sector-by-sector backup will occur, only sectors with data will be imaged...is this not true? I did the "sector-by-sector" backup, but did NOT have the box checked to include unallocated space. Given those conditions, how does Acronis create a 600GB backup file from a 223GB usage of the disk?

EDIT: Here is the section from Acronis Help that discusses the options I'm describing. I do not see why a sector-by-sector backup that excludes unallocated space should be almost triple in size from the Windows-reported used space size:

"Image creation mode
You can use these parameters to create an exact copy of your whole partitions or hard disks, and not only the sectors that contain data. Note that the Back up unallocated space check box is available only when the Back up sector-by-sector check box is selected.

To make a sector-by-sector backup, check the Back up sector-by-sector parameter. By default the program copies only the hard disk sectors that contain data. However, sometimes it might be useful to make a full sector-by-sector backup. For example, you have deleted some files by mistake and want to make a disk image before trying to undelete them, because sometimes undeleting may create problems in the file system. Please note that this mode increases processing time and usually results in a larger image file because it copies used and unused hard disk sectors.
The Back up unallocated space option becomes available if you have selected the previous parameter, Back up sector-by-sector. By default, while performing sector-by-sector backup, unallocated space is not included into the backup file. Enabling this option will include all unallocated disk space into the backup."

Unallocated space is the part of disk that contains no partitions. It is *not* sectors within the partition that are not used by file system. With sector-by-sector, it doesn't matter what sectors withing the partition are occupied by file system - all sectors of partition are backed up. For *disk* backup with *EDIT* 'backup unallocated space' /EDIT mode on, areas without partitions are also included.

Thanks for clarifying that, dev-anon, just goes to prove you can learn something new every day. Personally, I can't imagine having "unallocated space" on a hard drive that wasn't partitioned for use, and I also can't imagine it occurring often enough to need a tick box option for it in Acronis, but apparently it's there for a reason.

Ok, after further testing and diagnosis, I'm finding that my hard disk does have some sort of problem. I use Steve Gibson's SpinRite 6 utility and a few months ago, this drive had no issues. I just re-ran SpinRite (after a 2nd full run of "ChkDsk C: /r" found one file that was corrupt) and even SpinRite now crashes with an unrecoverable error.

I also retried the Acronis full backup process again with same params and I was able to get a much better backup WRITE process which ran continuously just like the validate READ process I originally posted. And...this time it was even faster, clocking a nice 70MB/s - 90MB/s all the way through the first hour or so, except for the place where I hit the sector read error again. Once I clicked "ignore" the backup resumed just as fast and continuously as it started. I'm happy that Acronis does seem to be quite capable of such high-speed writes that are of the same quality as the reads! Unfortunately, it appears I may have to be replacing my internal disk before I can do further testing to confirm. I will post more results once I line out the disk problems.

I've sent a tech support request to Gibson Research about the SpinRite crash, but I don't think it is necessarily due to my internal disk drive because the same crash now occurs on the new eSATA drive which I'm using as my backup target drive. Hopefully, I can get the SpinRite issue resolved and run a complete test on the internal SATA drive.

Meanwhile, I went ahead and did another full backup using data-only this time instead of sector-by-sector. I did not encounter the sector read error this time, so perhaps the disk error is on an unwritten sector. Again, Acronis performed VERY well, turning out continuous 70MB/s - 90MB/s write times on the backup (according to HD Tune Pro) and about 110MB/s sustained on the validation...awesome! (I think the OBSERVED disk write speeds reported from HD Tune Pro are a bit misleading due to the fact that they are second by second reports, whereas the ACTUAL calculated times are more accurate as they are based on total data volume divided by actual elapsed time. HD Tune doesn't give you an average, only current readouts and it's easy to overestimate.) I used normal compression, no encryption, and removed the "sector-by-sector" setting. The used data space reported by Win7 for my source drive was 229GB and the resulting archive turned out to be 166GB. The whole process, including validation took 73 min, with the backup function completing in 47 min and validation taking 26 min. So, the actual CALCULATED net average backup/validation speed turns out to be roughly 53MB/s! That is pretty SWEET!

Once that was done, I also ran an identical backup process from the same source disk to the WD MyBookWorld through the gigabit LAN connection for comparison with the eSATA hookup. I only let this backup run to about 30% completion, as it was only needed for a speed comparison. The backup disk write process held steadily around the 15MB/s mark and was clearly much slower than the eSATA target drive.

One last test for a reference comparison was a single 50 GB file using standard Windows copy/paste from the same source drive to both target drives. The 1TB WD MyBookWorld drive on gigabit LAN copied the 50GB file at a sustained speed of approximately 15MB/s and indicating a full hour+ to complete (I canceled it early). The eSATA connected Fantom drive copied the same file at a steady 65MB/s, completing in less than 13.5 minutes.

I'm done for now until I can diagnose and repair/replace my internal SATA drive, but I've learned a lot from this process and hopefully it is useful to others. Can't wait to try out the SATA III or USB 3.0 setups! My only complaint with Acronis is the near-useless backup time calculation function, but for raw performance, Acronis DELIVERS and I'm sold.....again!

Update:
My WD10EADS drive failed the WD LifeGuard extended diagnostics test and the sectors would not repair. I then restored the drive using the HP Factory disks and ran their diagnostics which also failed. But before filing a warranty claim, I decided to try a repartition of the drive and then restored my latest Acronis file backup (non-sector-by-sector) to see if that helped. Afterwards, I attempted the Acronis sector-by-sector backup again which is where I first found the problem. This time, it worked without sector read errors and it also verified. I then re-ran the WD LifeGuard utility...all tests passed without errors this time. Hmm... It seems a repartition and restore did the trick. I did not retry the SpinRite 6 diagnostics due to conversations I had with their tech support which indicated that the SpinRite failure codes were generated because the modern drives are just too fast for their software. They have had a message on the GRC site for the last four years about a possible SpinRite v6.1 update for SATA, but that doesn't seem likely to happen. Too bad, because their utility is otherwise very good.

Anyway, I've now run some additional file copy tests on the Fantom 2TB running on both the eSATA and USB 2.0 cable attachments. I will post those results later once I have time to go through the screenshots.

Ok, here is a summary of my unscientific results for the following drives and transfer situations/methods:

Source System/Drive: HP e9220y AMD Phenom II X4 8GB RAM on Win7 Pro running WD10EADS-65L5B1 - 1TB SATA internal 7200RPM
Target Setup 1: Fantom G-Force 2TB GF2000EU32 External (Samsung - Model HD204UI 7200RPM) on eSATA connection via Rosewill RC-219 Silicon Image PCI Express eSata x2 NCQ non-RAID SATA II Controller Card
Target Setup 2: Fantom G-Force 2TB GF2000EU32 External (Samsung - Model HD204UI 7200RPM) on USB 2.0 conenction
Target Setup 3: WD MyBookWorld White 1TB NAS tested on 1 Gigabit LAN

Acronis TIH2011 build 6597 was used in Windows 7 with normal compression, no encryption, with validation, and no sector-by-sector. Windows reported the used portion of C:\ as 229GB

RESULTS
Acronis Backup Archive Process from Source Drive to:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 53MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 25MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 15MB/s

50GB Single File Copy/Paste:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 65MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 30MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 15MB/s

47.5GB Folder 22K+ various files:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 37MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 16MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 7MB/s

My interpretation is that a single large file seems to have the best overall read/write performance times, but I'm VERY impressed with Acronis' ability to keep up with the Win 7 standard file copy/paste times across any of the three connections. It looks like everything slows down considerably when multiple file types and sizes are in the mix.

I hope you find this information useful. I also hope to test USB 3.0 and SATA III connections at some point in the future.

Just wanted to add that I have installed a PCMCIA Express Card eSATA controller device to my wife's HP Vista Home Premium powered laptop (Pavillion dv6700 3GB RAM) which has made a HUGE difference in backup times for me. I had been getting very dismal backup times across our wireless network (no surprise). I had even tried adding a USB to 1 GB LAN connector to speed things up, but was still only getting maybe 3MB/s - 5MB/s speeds with backups still taking several hours to complete and verify.

I found a very inexpensive PNY brand eSATA adapter online which really works quite well. I first did an assorted file copy transfer which was still very slow in the 5MB/s range to the Fantom eSATA drive. However, when I did the Acronis backup, I found I was getting 32MB/s average NET write/read and the 75GB backup completed and verified in just 40 minutes. Very acceptable...

I just came across a couple of free tools that may be of use to someone. I like HD Tune Pro, which I mentioned above, to monitor disk read/write speeds during actual backup operations. However, Physical Disk Speed Monitor from http://www.pcwintech.com/shanes-toolbox is a nice, simple, freeware disk read/write monitor without all the frills. You'll find the download link about 3rd from the top of the page. This monitor will allow you to see the real-time disk read/write speeds while other programs are using the disk, such as Acronis or a Windows file copy/paste function. There is no installation, only a .exe file. There is no graph of read/write activity, only real-time digital information.

The second tool is called Parkdale SpeedTest v2.68 and it is also a non-installation .exe file. This program seems to have three modes of operation, QuickAccess, FileAccess, and BlockAccess. I have not tried the more complex tests, as the QuickAccess function gives me the data I want. This program is also freeware. Here is the URL for Parkdale:
http://run.to/sz

You will find the program listed in the menu on the left side of the page.

Mr Pisky wrote:

Ok, here is a summary of my unscientific results for the following drives and transfer situations/methods:

Source System/Drive: HP e9220y AMD Phenom II X4 8GB RAM on Win7 Pro running WD10EADS-65L5B1 - 1TB SATA internal 7200RPM
Target Setup 1: Fantom G-Force 2TB GF2000EU32 External (Samsung - Model HD204UI 7200RPM) on eSATA connection via Rosewill RC-219 Silicon Image PCI Express eSata x2 NCQ non-RAID SATA II Controller Card
Target Setup 2: Fantom G-Force 2TB GF2000EU32 External (Samsung - Model HD204UI 7200RPM) on USB 2.0 conenction
Target Setup 3: WD MyBookWorld White 1TB NAS tested on 1 Gigabit LAN

Acronis TIH2011 build 6597 was used in Windows 7 with normal compression, no encryption, with validation, and no sector-by-sector. Windows reported the used portion of C:\ as 229GB

RESULTS
Acronis Backup Archive Process from Source Drive to:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 53MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 25MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 15MB/s

50GB Single File Copy/Paste:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 65MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 30MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 15MB/s

47.5GB Folder 22K+ various files:
Net Sustained Transfer Rate to Target Setup 1 (eSATA) : 37MB/s
Net Sustained Transfer Rate to Target Setup 2 (USB2.0) : 16MB/s
Net Sustained Transfer Rate to Target Setup 3 (1GB LAN) : 7MB/s

My interpretation is that a single large file seems to have the best overall read/write performance times, but I'm VERY impressed with Acronis' ability to keep up with the Win 7 standard file copy/paste times across any of the three connections. It looks like everything slows down considerably when multiple file types and sizes are in the mix.

I hope you find this information useful. I also hope to test USB 3.0 and SATA III connections at some point in the future.

Just built a new system based on an ASRock 890FX Deluxe5 mobo with full SATA3/USB3.0 support. I have an OCZ Solid3 SSD SATA3 boot drive and a WD Caviar Black 1TB 64MB cache spinner for data (will RAID it later). Initial full Acronis backup of 177GB source data from the OCZ/WD Caviar combo transferred to my Fantom eSATA 2TB target drive at sustained 100-125MB/s speeds. That's double what I was getting from the SATA2 hardware. I'm no longer using the Rosewill PCIe eSATA card since this mobo has eSATA on the rear I/O panel.

I don't have any USB3.0 devices to try out yet, but maybe soon. Will probably get a small travel disk with USB3.0 if my wife upgrades her laptop.

Sorry for crashing the thread, but it seems to have been abandoned. To answer the OP's question, I guess I'd have to say ditch network transfer methods and migrate to faster hardware platforms that support SATA3 and USB3.0. Acronis does a very fine job when operating on capable hardware. I could just never get my networked target devices to write data fast enough to be useful. Hope someone finds this info useful.

I find it very helpful. Thanks for sharing.

Over the years I've seen the odd thread that seems to indicate very good network speeds but I always regarded them with suspicion. I no longer consider using a GB LAN as a preferred mechanism for doing imaging because of the speed limitations and typical large sizes of images these days. I always had a unscientific feeling that a GB LAN was about the same as using USB2.

Imaging speeds have indeed increased a lot over the last few years. Earlier my estimate on a typical good (but not the best) PC was to estimate about 1 minute/GB of normally compressed archive. On a modern good machine I'd easily say about 0.3 to 0.5 minute/GB of the same data. While disk/network speeds can limit image throughput, a slow processor can also impact the speed especially if any compression is being done.

I bought a PCIx Vantec USB3 card and a USB3 enclosure for my 3.2 GHz P4. A quick test showed the backup time to be the same as using an internal SATA2 drive. The drive in the enclosure was a SATA2. I was happy since that is a big improvement over USB2.

In the posts above I would not be satisfied with the 15 MB/s and 7 MB/s transfer rates across a GB LAN - there must be something wrong. My experience has been that a backup to a disk across a GB LAN is limited in speed by the transfer rate of the disk (source disk OR destination disk).

The transfer rate of the disk drive on my network's server is 60 MB/s, so images that are saved across the GB LAN to the server proceed at the transfer rate of the slowest disk involved in making the image; typically 60 MB/s. Images saved to another disk on the same PC proceed at 60 MB/sec also - that's the fastest rate that my spinning hard disks are capable of. The laptop's SSD, however, is capable of 120 MB/s transfer rates when imaging to a partition on the same disk, but limited to 60 MB/s when saving the image to a spinning hard disk. In all cases it's the disk's transfer rate that's the limiting factor; not the GB LAN.

Mark Wharton wrote:

In the posts above I would not be satisfied with the 15 MB/s and 7 MB/s transfer rates across a GB LAN - there must be something wrong. My experience has been that a backup to a disk across a GB LAN is limited in speed by the transfer rate of the disk (source disk OR destination disk).

The transfer rate of the disk drive on my network's server is 60 MB/s, so images that are saved across the GB LAN to the server proceed at the transfer rate of the slowest disk involved in making the image; typically 60 MB/s. Images saved to another disk on the same PC proceed at 60 MB/sec also - that's the fastest rate that my spinning hard disks are capable of. The laptop's SSD, however, is capable of 120 MB/s transfer rates when imaging to a partition on the same disk, but limited to 60 MB/s when saving the image to a spinning hard disk. In all cases it's the disk's transfer rate that's the limiting factor; not the GB LAN.

Good comments, Mark, and I agree. I probably mis-worded my posts to give the impression that I was holding the LAN responsible for the slow times. I believe in my situation the slow "network" speeds were related to a bottleneck at the WD MyBook drive, either its ethernet card, the disk itself, or a combination. I have not taken the opportunity to test backing up to a network share on my old SATA2 HP which would give me an alternative SATA3 > 1G LAN > SATA2 setup that will likely outperform the MyBook. I'll post when I get the results.

Regarding processor speeds, I can attest to the fact that my current Phenom II X6 1100T blows away my previous Phenom II X4 910 hands down. The Phenom II X6 ran Memtest86+ v4.20 almost four times as fast as the Phenom II X4 during overnight runs on the exact same set of 16GB Gskill RAM chips. In the last couple of years, I have used and tested Acronis on Toshiba, Dell, and HP laptops and my 2009-2010 model HP Elite desktop and finally my latest build on the new SATA3/USB3.0 AMD 890FX platform with SSD. I continue to be amazed at Acronis' ability to keep pace with the latest advances and data throughput capability.

Seekforever wrote:

Over the years I've seen the odd thread that seems to indicate very good network speeds but I always regarded them with suspicion. I no longer consider using a GB LAN as a preferred mechanism for doing imaging because of the speed limitations and typical large sizes of images these days. I always had a unscientific feeling that a GB LAN was about the same as using USB2.

My experiences with GB LAN have left me with the feeling there was some potential there for greater speeds, but in all my attempts to find bottlenecks I always found that the endpoint hardware was keeping me from seeing those blazing speeds I heard others speak of. I deal with at least four home machines that must be backed up regularly and 8 hours to run a backup program across a network just isn't acceptable. But as Mark has already pointed out, the slowest link in the chain is always going to hold everything back. Ideally, I'd like everything to be networked in the house and have everything running at a minimum of 50MB/s. If I could get that, I'd probably be happy with my LAN.

Well, I just had to try this out, thanks to all your comments and I'm glad I did too.

I just re-checked my wired network with Ixia's QCheck endpoint system and I'm getting between 500-800Mbps (62-100MB/s) TCP throughput on the LAN (UDP is around 180-200Mbps). It's been a while since I checked it, but I've never gotten actual devices to run data that fast over the LAN. So, I setup a full backup of my new system, currently about 180GB between the SSD and the data drive. I targeted the HP Elite's native SATA2 WD10EADS 1TB drive to receive the file. Watching it real-time showed varying MB/s speeds anywhere from 70-120MB/s. Acronis reported a start to finish time of 66 minutes which calculates to 45MB/s for 180GB. Although the result is good, I was surprised at the 45MB/s figure because it really looked to be about 80MB/s or MORE watching it live on HD Tune. Maybe my calculations are wrong somewhere?

Either way, that's a SERIOUS LAN improvement from backing up to the WD MyBook, which is also a WD10EADS SATA2. It also proves Mark's point perfectly. Incidentally, I just found out that while the MyBookWorld 1TB drive has a 7200 RPM drive, it also has a slow 200Mhz processor which accounts for the very slow transfer rates.

Mr Pisky wrote:

Just wanted to add that I have installed a PCMCIA Express Card eSATA controller device to my wife's HP Vista Home Premium powered laptop (Pavillion dv6700 3GB RAM) which has made a HUGE difference in backup times for me. I had been getting very dismal backup times across our wireless network (no surprise). I had even tried adding a USB to 1 GB LAN connector to speed things up, but was still only getting maybe 3MB/s - 5MB/s speeds with backups still taking several hours to complete and verify.

I found a very inexpensive PNY brand eSATA adapter online which really works quite well. I first did an assorted file copy transfer which was still very slow in the 5MB/s range to the Fantom eSATA drive. However, when I did the Acronis backup, I found I was getting 32MB/s average NET write/read and the 75GB backup completed and verified in just 40 minutes. Very acceptable...

Just to further solidify my belief that HARDWARE limitations are the most likely and most significant cause of backup/recovery speed bottlenecks, the 200GB Seagate Momentus 5400 RPM hard disk in my wife's HP Pavillion dv6700 laptop died last month. It was replaced by a WD Caviar Black 320 GB 7200 RPM unit. The Seagate drive had been previously producing about 32MB/s throughput on Acronis backup jobs, but the new WD Caviar Black has now bumped up backup speeds for this computer to a solid +/- 50MB/s. Now...if I could just get that to pump through a wireless network at 50MB/s...I'd be in high cotton! I can DREAM can't I? LOL

I still haven't had the budget to try any USB 3.0 devices, but I'm sure they are going to be fairly similar to the SATA3 interface. My current desktop build which uses a WD Caviar Black 1TB SATA3 internal HD combo'd with an OCZ Solid3 SSD boot drive is still cranking out +100MB/s backups to my target 2TB Fantom eSATA external storage drive. I have not yet had any need or desire to upgrade my current Acronis TIH 2011 build 6942.

Excellent results - I'm glad things are working faster for you. Perhaps soon you'll be able to get faster wireless speeds than the Wireless G standard of 22 Mbps, which is slightly less than 3 MBps; we only need about a 36x improvement to get to 100 MBps.