djg05, if all you're interested in is FULL backup speeds, you won't find much difference... but the new RESTORE SPEED and its SCHEDULING CAPABILITY are close to groundbreaking. PS- the download links HERE always pick up the latest pre-Releases.
In practical terms, suppose that I create a full image and then, immediately, restore it. Does this "groundbreaking" speed mean that this restore is done in (essentially) zero time?
Yes. Of course "essentially" zero time could be anything from NONE to hours, I guess. I've never seen "essentially zero time" actually defined anywhere so I'm not sure what that means. Groundbreaking? Yes, for Macrium.
I just, against my better judgement (related to testing unfinished products), tried out the v6 trial. Created a Full Backup of my System Drive, "Immediately" created an incremental of the System Drive and then created a rescue boot media (usb thumbdrive) booted into the rescue media and restored the Full backup (with Rapid delta restore enabled) and also chose to verify it. Took 3 minutes XX seconds to restore of which 2 minutes and YY seconds was the verification stage. I believe the actual restore was less than a minute. That's using two SSDs. Oh and it seems to have finished with no issues =) This may be my backup program in the future when it releases, I'm not so excited about buying a license at the moment when I don't know what the V6 license will look like/cost... Actually I can't afford it at the moment anyway so.. yeah. When making the definition (if I'm using the right terminology here, new to this product) I noticed that I could choose to retain X amount of Full, Differential and Incremental backups. I assume that they merge when the number has been reached? My question is if there is some way to set that number to infinity? I'd like to use it like I use AX64 at the moment, which is one Full backup and then incremental on incremental for infinity - All manual of course. I guess I could set the incremental number to something ridiculous but that.. I mean.. I just.. There's something inside me that just tells me I don't want to do that because it like looks weird or something........ Edit: I'm talking about the "Define retention rules" Edit: Is it as simple as unticking them? Edit: I think so?
This is what I was thinking about. If no changes have been made, the restore time should be very short. Zero, in theory, not realizable in practice.
From the tests that I run the minimum restore time seems to be similar (a bit longer) to the time needed on your system to perform a CHKDSK /F on that partition. Panagiotis
Froggie, thank you, for asking them and reporting back. I agree with you. Macrium 6 seems very robust from the so far testing that I gave it (keeping the incrementals to a max of 10). I hope to find time this week to do some stress testing e.g. how it behaves when taking an incremental during high I/O write activity on the partition. (most hot imagers fail this test) Panagiotis
Going back to David's question about how times compare with Active@disk. The system this was tested on has a 124gb c: drive and the imaging target was another internal disk drive. Full Image time Active@disk 14:32 Macrium 13:13 Incremental image time Active@disk 14:28 Macrium :55 to 1:05 including merge Restore time Active@disk 21:26 Macrium 5:56 Pete
these are very impressive figures - it is likely that Macrium will knock off Shadowprotect in your arsenal , Pete?
Hi Robin! I timed a 35.3gB process like you described from the SHUTDOWN desktop on the way to the automatic WARM restore to the standard DESKTOP upon its return... the times include the reBOOT, the BOOT into the RECOVERY PE, the full recovery and the reBOOT back to the system and all those silly timers the process gives you to change your mind along the way. 2m 55sec to do the complete restore including all the times above... FYI. The actual restore process in the PE does nothing other then compare the current file structure on disk to the file structure contained in the image... then using that info, scans the image (and it's not really a scan, the image appears to have a container format that is indexable... the scan occurs too fast to be a real image scan) and replaces only the blocks that have changed since the image was taken... that part of the process is really pretty quick, probably much less than a minute. Most of the time is in the shutdown, BOOT in to the PE then reBOOT into the system.
For those folks wondering what the differences between the new HOME and WORKSTATION versions are... NONE. Feature-wise they will be identical. The difference will be the SUPPORT MODEL used for each... it will be very different for HOME users than WORKSTATION users. I do not know the support differences, BUT... a new v6 website should be available in "the coming days." And no, I have no idea if there will even be a FREE version, and if so, what features may be absent that version... sorry. Let's wait for the website.
Given the portion of the v6 manual I quoted (and you provided the URL), I think there will be a FREE edition. ""Macrium ReDeploy is now included all editions of Macrium Reflect except for the Free Edition."
The idea of I/O stress testing was interesting so I thought about what kind of test I could do here. I decided what better way for massive I/O activity then defragging. So, I started a defrag with Perfect Disk. Once it was defragging I ran an Incremental backup. It ran ran fine, just took longer. When it was done, I cancelled the defrag and tested the image. It seemed fine so I did a restore. It was perfect. Anyone else have ideas for stress testing? Pete
What you're describing is more of a test of Windows and VSS. Macrium doesn't handle I/Os. Windows does.
True, but I think what the Froggie was curious about was how Macrium would do under those circumstances.
Hi Pete, defragging is good for stress testing, but do not use PerfectDisk (it has I/O throttling technology and slows down in such situations), use mydefrag instead (especially the script "Move all the files and directories to the end of the disk" is great for this test). And in Macrium enable the feature to check the image after is taken. Correct, but is Macrium's responsibility to identify if something went wrong during the imaging process, eliminate the incremental taken and reschedule it. Panagiotis
I have found that it is the backup speeds that have been the most time consuming with Macrium 5. I use it for the data partitions on differential. These only need to be done weekly. I was using the system partition as a way of comparing the two. I have just tried one of my data drives that has 58 GB on it - Active 6, Macrium 10 with Active creating a slightly smaller file. I suppose it depends on whether you do a lot of restoring or not. Thanks for the link - I got 469
Thanks for trying that out Pete. Interesting that you get the opposite results to mine. See my reply to Froggie on the data drive test . I can certainly see that restoring the sytem drive does not need to be done from cold, if I have understood it correctly, and that would save a lot of time. I do that more often than any other restore.