Cold-Imaging Competition

Could you do the same test to ShadowProtect?

 

I would have tested SP if I had a copy.

Aaron

 

With a trial is enough? you can request it in their website

 

I believe to do cold imaging you need to fill out a special form to get a copy that allows use of the boot disc. I filled it out twice ages ago and never heard back. I bought it at one point though.

 

I had a similar result when I requested the SP boot disk and then when I learned that SP had no provision for generating a custom boot disk with special drivers (as with ADI and AIP), I didn't pursue it any further.

Aaron

 

Hey Aaron,

Your test results sure seems to speak volumes for AIP so I just downloaded the trial and will give it a go...

Thanks for all your effort!

Scott

 

Hi Scott,

I'm glad you are going to trial AIP before buying it. There's no question that in the couple of chances AIP had here it ran like a champ, but the jury is still out when it comes to whether or not it can do that consistently! Let us know how it's doing the job for you.

Aaron

 

I've been using AIP for about a month now and doing many restores and it's been 100%. Again, my needs have been basic...a weekly cold full image. I've done 1-1 restores and 1-? restores (where I change the partition size). I think it will only get better, and I would advise anyone to download the free trial if interested. It's still not my primary imager at this point.

 

To contribute some numbers, I timed two recent operations I performed, using Paragon HDM 2011 Suite from a WinPE USB key and a ShadowProtect WinPE 4.0.1 CD. I created images of the 27 GB system partition (Windows 7 x64) in an external USB 3.0 disk connected to a USB 3.0 port, and then I verified it.

To create the images, I used the default options with image splitting set to 4 GB. In HDM, defaults include "Control archive integrity", which "can slow down the backup operation".

The results:

Time to boot (measured from clicking on "reboot" to display of main program window): HDM 2:10 min, SP 2:02 min.

Time to create the image: HDM 6:05 min (2 min "preparing" the operation and displaying somewhat idiot messages), SP 4:09 min.

Time to verify: HDM 2:02 min, SP 1:42 min.

Size of the images: HDM 9.8 GB (about 51% compression excluding hibernation and page files), SP 9.3 GB ( 53.5% compression excluding the same files).

 

Thanks Robin A....I have used both and find Paragon to always be the slowest. I am not sure why that is. On the bright side it's always been stable and you can tap into all the other partition management features before and after imaging. SP is hands down faster for me as well.

 

napoleon, have you timed AIP vs. SP (for cold-imaging)? Several members asked me to test SP but I don't have the boot disk so perhaps you can add some insight as to how well it performs compared to AIP (this competition's champ).

Aaron

 

The "check integrity" option will be slowed down the speed a bit.
Maybe you can uncheck it(I find there are two integrity options) and test again.

Generally, I don't use the default options in PHDM or in other imaging apps.

 

Good point. If I can carve out some time within the next day or two I can easily do this on my laptop, which never takes too long to image. If I can I will post the AIP vs. SP results here. Thanks.

 

With all the talk of running Active@ Disc Image from the Active@ Boot Disc interface, I went for it and bought one. In the past, I have used the password reset program they make (saved me big time!!) and also felt the file recovery portion would someday pay for itself. So buying the whole thing was a no brainer for me.

I have now used it on my netbook and it has done a good job. Backups and restores went quickly. One issue though. If you drive has more than one partition, it appears you MUST backup ALL partitions and restore ALL partitions. This is because when ADI restores, it wipes the target drive completely. If you only backup, say the C drive, and then restore only the C drive, that's all you'll have when finished. Any other partitions will be gone. This is very different than most other imaging apps I have ever used. Others I have used allow you to backup and restore each partition individually as you desire.

Again, the program seems rock solid. But if you have a multi-partition drive, save and restore them ALL! Otherwise you may get a an unhappy surprise!

 

Speed tests for backup programs reminds me of speed tests for video cards when looking at something for a HTPC. It's almost irrelevant. Compatibility and reliability are more important for the former, while video quality more important for the latter.

If you can test a restore with different problematic chipsets from a restore media, like with NVidia or VIA chipsets, different IDE/SATA chips (Jmicron crap, for instance) and the like, then that is a true test. IIRC, Acronis failed miserably in the 9.0 and 10.0 days unless you built your own WinPE media. Maybe things are different today.

Being able to restore around 'interesting' partition tables and multiple OS's is also a plus. I know that Terabyte's backup programs do well in all these scenarios except resizing a partition smaller. Acronis for instance, does well in that area.

 

That "issue" would be a huge one. But the information available in the web page suggest that partition-level operations are possible and supported.

 

you

Hi valnar,

No one here, including myself, denies that compatibility and reliability are paramount backup factors (e.g., ATi refused to restore its image in my last test so it's speed advantage was meaningless). However, to suggest that speed is irrelevant is simply not true for cold-imaging.

I would certainly agree with your statement if we were discussing hot-imaging, where you can multi-task (run other apps concurrent with your backup). But with cold-imaging the PC is dedicated to the backup process, so the faster it gets done the better (as long as reliability is not compromized)!

Aaron

 

Actually, that doesn't change my complaint since a full OS restore by definition is a cold image. If you have hardware that works with your imaging program, then yes, speed may matter. If that is the only thing you do with the imaging program, than compatibility may be moot if you deal with the same hardware day in and out. But if you need a broad range of compatibility, some programs certainly fall short.

But yes, probably off topic. I'll agree.

 

They are. But only if you have a single partition drive or are willing to go from a multi-partition drive to a single. As I noted, the reason is because the program wipes the entire target drive before it restores anything. As an eg., you can't restore the C partition, then later decide to restore an additional one you backed up earlier. The second restore would blast the first one before proceeding with the 2nd one. Other imaging apps I have used do not do this.

While this sounds bad for anyone wanting a more advanced approach, it actually would probably be better for a less knowledgeable user. They would image the whole drive and restore it that way too. Simple.

 

@Aaron

Nice test but is a bit unfair, since you tested different enviroments for each application. For proper results you should test all those apps in the same enviroment and not in a mixture of DOS, Linux, WinPe and BartPE.

Dos performs very slow since it uses the bios to control the devices, both dos and bartPE do not support multithread apps, and the latest linux kernels perform much faster than the WinPE 3.x:

DOS= the slowest because of the slow transfer rate of the controller (since it uses the bios) and because it does not support multiple cores.
BartPE= must faster than dos because windows drivers perform faster on the same controller.(does not support multiple cores)
WinPE= Faster than BartPe because it can use multiple cores and multithread applications can compress the images faster.
Linux (with kernel 2.6.29 and higher)= have higher speed rates from the winPE3.0 or winPE3.1.

Bottomline your results do not reflect the speed of the programs but the speed of the OSes they operate on.

Panagiotis

 

The latest release now does that very well...

 

i can test sp, macrium, active@, ifw, ati and others as well if needed

 

Hi Panagiotis,

Your points are well taken, but contrary to your bottomline conclusion my results didn't entirely reflect the speed of the underlying OS (although I'm sure that played a role).

As can be seen below, in my first cold-imaging test (Windows Used Sectors), the results are pretty much in tune with what you are saying, but still shows clear differences between imaging programs using the same OS! ...while in my second cold-imaging test (All Sectors On Partition), there's no denying that AIP (WinPE) 'ran away' from the others, but what's really interestingly is that DS (DOS) easily beat-out ADI (WinPE)!

Test #1 (Windows Used Sectors Only):
1. ATI (19’ 40”) - Linux
2. AIP (20’ 00”) - WinPE
3. ADI (23’ 57”) - WinPE
4. DS (24’ 15”) - DOS
5. IFD (28’ 20”) - DOS

Test #2 (All Sectors On Partition):
1. AIP (26’ 50”) - WinPE
2. DS (30’ 20”) - DOS
3. ADI (33’ 10”) - WinPE
4. ATI (Refused to restore 'insufficient space on partition' error?) - Linux

In my defense, I tested programs I had onhand with their very own boot disks (I considered that a fair approach). The impact of CPU cores was of no consequence because my test PC has a single core Pentium.

Regards,
Aaron

-----
PS. I might add that I could have used a WinPE boot disk with DS (the method I normally use, which surely would have elevated DS in the final results!!!), but with fairness in mind, I used the boot disks as supplied with each program.

 

not true, i have 3 partitions on my c drive and i can restore each one individually or together, not sure what your doing that would deny you that ability.

i just restored 1 partition of the 3 that are on the comp i am writing this from and it did not wipe the other partitions

 

exactly