Impressions of TI 11

i recently purchased TI 11 for an as yet unpuchased/to be built Windows XP or Windows Vista system.

However, since I currently use a multiboot Wib 2000 system, wit TI 9 install only in the main OS, I figured that i'd temporarily install TI 11 in one of the other OS on the same PC.

1. As I indicated in another thread, tho I could boot from the retail CD, the AutoRun did not work. I had to navigate to the Setup.exe in theTrue Image directory to install.

2. I had hoped that the install would provide options to not install particular features, such as Try and Decide. Did I miss the options during the install.

3. I then ran a full image backup of all my internal drives (3 SCSI drives with 10 logical drives).

On 17 Nov 2007, I had run such a backup with TI 9: Backup used 2:30:45, Verify used 43:07.

On 3 Dec 2007, TI 11 Backup used 4:16:24, Verify used 1:15:18.

Very poor results.

I've uninstalled TI 11.
Hope that it performs better on whatever system I next use.

Note that I used buid 11.0.8022. No point in downloading the latest build until I am ready to install on a system fer good.

 

Build 8,022 is the retail release and probably even more buggy than 8,027. The current build (8,053) has a lot of fixes in it.

I disagree that there is no point in installing the latest build. If you want to test it out, then test with the latest build. There is no point wasting time on an older build of TI 11.

When you ran your backup comparison test, was it from Windows or from the TI CDs?

Was it internal to internal or internal to external (USB, etc.)?

 

from windows.

i would never backup to internal.

 

Your times (even for TI 9) seem odd. Why does the backup take so much longer than the verify? Most times the USB speed is the limiting factor and the backup and verify times should be almost the same. Or is this a NAS or network share backup?

When I backup 20GB to a USB hard drive it may take about 20 minutes (approx. 1GB/min.). The verification takes the same amount of time. On internals, I get 3-4GB/min. on both backups and verifications.

 

This thread is pointing out that v 11 is much slower than v 9 on a system with all SCSI internal drives and USB 2 external drives using the same hardware.

Why this system may be slower than others is dependent on the speed of the drives, memory, etc, but them's not the point here.

 

I wasn't trying to directly compare my results with yours. I was just pointing out that even with TI 9, the backup is taking about 3.5 times longer than the verify (a ratio that holds for TI 11). This either means the computer has a slow CPU and compression is taking a long time or there is something else going on with drivers, etc. since I assume your SCSI drives are fast.

Others have also reported that TI 11 takes longer than previous versions, though I haven't noticed it myself.

If you don't mind, could you post the following information:
- What build of TI 9 you used
- The total amount of GBs backed up from the selected drives
- The finished image size for each backup
- What compression level was selected

If someone else posts that uses SCSI drives, it would give them something to compare to.

 

A verify MUST be much faster than a backup.
In any case, my only concern was how much slower TI 11 is than TI 9. on the same PC.

Only difference is that TI 9 and TI 11 are installed in diferent OS partitions (both same update level of Win 2000). For th epurposes of comparison, the othyer differences between the partitions should not matter. And same backup drive was used with plenty of space, so fragmentation should not be an issue.

 

This thread once again points out the need for a throughput indicator, so we could see exactly what level of performance is being achieved, and isolate the bottlenecks. I've suggested this feature since version 6, to no avail.

 

A throuput indicator would not offer any additional info.
The throughut can be calculateed by using the times in the log, if you know how many bytes are being backed up.

You can use GetDiskSpaceUsed to learn th source drive sizes.

 

I do not knowhow many bytes were on the source drives either on 17 or 3 Dec 2007, but on 4 Dec 2007, there are about .5GB LESS than on 9 Nov 2007.

Both backups ran with Normal priority and Compression. AFAIK, no other options differed.

My recollection is that the 3 Dec image was about 100MB larger than the 17 Nov image.

I've already uninstalled TI 11 and deleted the .tib files, so that's best I can do.

 

I disagree with this. On my computers, the backups and verifications take approximately the same amount of time. With a fast CPU, the bottleneck is the disk throughput, either USB 2 or SATA-II.

Understood and I agree it with you. However, TI 11 uses a different backup format and that's probably what causes it to be slower on some systems. Since your backups were from Windows using the same Windows drivers, the only difference was the TI versions and their internal processing.

You say that TI 9 took 2.5hrs to backup the image. How big was the final image? 50GB, 100GB?

 

I recently bought TI11 and installed it on my Laptop (it is build 8022). I thought about upgrading to the latest build but there is no patch. You have to download the whole product again. I have a dialup connection and its a 137 Mb file. Isnt there a patch available. Not everyone can afford DSL.

 

That's the way Acronis puts out "updates". They're the complete program. Can you go to your local library and download using their wireless system? Or perhaps any nearby hotel that has free wireless would work?

 

Regretably no I can not. Thanks for the suggestions though. Perhaps if they would let me purchase a later version burnt onto CD that would be ok. Something for them to consider at least.

 

Do you mean "USB or Esata" or "USB or Sata 3.0"? I would not expect Sata 3.0 to be an issue -- it would be at least as fast as Sata 1.5 or or the earlier IDEs, no? I would guess that in virtually every case the drives are the "bottleneck" but Sata 3.0 wouldn't be especially slow compared to other drives, would it?

Or are you saying that ATI can't handle Sata 3.0 properly?

 

I disagree. A throughput indicator would provide a REAL TIME indication of how fast the transfer was. (Without having to use a calculator, or view the log file.) Back in the old DOS Ghost days, I used to image many systems for a client of mine. (Yes, I had a Ghost technician license, so this was perfectly legit.) I would cue up each machine, and start them one at a time. When I hit the point of saturating the network, ALL of the systems would show a drastic decrease in throughput. I would cancel the imaging on one system, and the throughput would jump up to normal levels. (At that point, I knew exactly how many systems I could image on the customer's network.) Since every network is different, a real-time throughput indicator is the only way to do this type of analysis. In addition, a throughput indicator allows me to tell at a glance whether I am running on a 10MB, 100MB, or Gigabit network, just based on the throughput.

 

shieber,

What I meant was that on my computers the bottleneck is my USB 2 hard drive or my internal SATA-II (SATA 3Gbs) drives. There are no problems with either. I was just trying to say that the CPU could handle the compression and backup fast enough that the only thing slowing it down was the throughput to the backup device. I usually get about 1-2GB/min. in USB 2 and 3-4GB/min. on SATA-II. This is consistent for backups, validations and restores. If I were to backup a RAID 0 to another RAID 0, I would get an approx. 7-8GB/min. bottleneck. (These speeds on my newer computers with E6600 and E6750 CPUs.).

In Howard's case, his backup times are 3.5 times longer than the verification. That means that something else besides the disk throughput is causing the slowdown. I would assume that his USB drive is not the cause because he can verify fast so the problem must be with how TI is reading from the SCSI drives or the CPU is slow on the compression. He never posted the finished image size (understandable, since it has been deleted) or his computer specifications so there is nothing to calculate or compare to.