Macrium Reflect

I stand corrected on "B" above... you cannot use the same folder for different image sets. The spec for setting them up sounds like you can (Imaging to different image sets) but when they are in the same folder, scheduled operations don't always work against the image set they're supposed to. I discovered this whle doing the above.

I think it's a BUG, but since we're not "ofishuly" TRIALing or BETA testing, there's no way to report it Anyway, the scheme works fine when the chains are in different folders.

 

The idea, of course, was an AX64 that was reliable. I wonder if Isso had remained in charge if that would be the case by now. It looked like it was close with the last beta and then things fell apart big time. Real shame IMO.

 

Thanks for your answers and help TheRollbackFrog and Peter2150!
Edit: one more little thing i need to know,is there a way to write a description comment in the first Full Backup for example (Baseline) and in the incrementals for example (Before Windows Update February) and so on,like i do in Image for Windows?

 

Thanks for your post - this is a great feature - I found that with a long chain AX64 could take 2.5-3x longer to do a delta restore than it would take to do a full one.

 

Actually, it is the opposite. IFW "Write Changed Sectors Only" feature is faster than a normal restore (most disks have higher read speed than write spead). On my disks usually is 20% faster.

It compares only the sectors that are to be restored, not the entire surface. True IFW incrementals are slower because of this comparison but it is the only way that guarantees that the image taken is an exact copy of the disk (even if a sector was modified with direct disk access).

Panagiotis

 

You are welcome.
I have not tested chains longer than 10 incrementals though...

Panagiotis

 

I'm planning on testing a chain of 30 tomorrow - it won't be the greatest test as all the incrementals will be from the last three days, so no large merged incrementals in there (not even sure if Macrium can merge incrementals (apart from with the baseline)).

 

Yes, but it's not quite as easy/straight forward as others. When the REFERENCE (the first) backup is done and all the specs are set, one of them is a COMMENT field. When the spec is done, it will ask you if you want to run the request AND save a copy of those specs for future use. You should do both. The backup will run and a new DEFINITION will appear for future use (so you don't have to give it all that information again). If your 1st operation was a FULL backup, when you get ready to do your next incremental, the DEFINITION spec in Macrium can be <right-click>ed and "Advanced options" selected, which will allow the change of some of those specs... one will be the comment. Change it, then run the DEFINITION, select INCREMENTAL, and the incremental will run with the new comment. This comment will stick with the DEF file so if you want different comments each time you use the DEF function, you'll need to make the comment change.

 

Sure, but for IFW on a TRIMmed SSD, the restore operation will have to consist of a READ followed by a WRITE for every sector contained in a file to be restored... a normal restoration would use just a WRITE. For an HDD there would be quite an advantage, I would think.

Am I missing something here...?

 

Chamlin, I have your operating environment set up as I type. I have a pinned (START menu) icon which does nothing more than a manual incremental snapshot when needed, and yesterday, I started what I'm calling TARANTULA MODE (the crawling baseline). This one saves 36-snaps (incl. BASELINE which is really Incremental #1) at a time and will constantly create what Macrium calls a SINGLE SYNTHETIC DELTA (a moving baseline which merges into the oldest incremental of the limited 36-count chain). For me this represents about 1-1/2 days of system operating experience... you can make it as long as you'd like.

Along the way while in TARANTULA MODE I'll be doing various restores just to see how 36-incrementals affect typical restore times... wish me luck!

 

Some delta restore timings with a baseline and 20 incrementals (averaging about 100mb each).

Current state > baseline: 2m16
Baseline > incremental #20: 2m07
Current state > incremental #18: 2m03
Incremental #18 > # 20: 1m36

 

Well... maybe TARANTULA MODE is not quite what happens. What I actually see is that the changes in the oldest INCREMENTAL (not the BASELINE) are merged into the BASELINE and that incremental becomes no longer. So what appears to really be happening is that the baseline remains at its place in the time line and each successive SYNTHETIC DELTA brings the next incremental in the timeline back into the baseline. So the only "crawling" part are the 36-incrementals themselves... the baseline remains in its time spot.

Will rethink this a bit...

 

A check, via Windows Explorer, says that the Creation/Modification dates of that BASELINE file are still those of the original... BUT, when looking at the chain in the Macrium RESTORE window, the date/time of the file that now has the BASELINE's original name is the same as the first incremental that was "merged."

Based on this info, I really can't tell whether the "merge" operation was done BACKWARD into the original BASELINE or FORWARD into the original INCREMENTAL #1. The main difference would be the point-in-time that this file actually now represents.

Does anybody have an answer to this question? Is this really TARANTULA MODE or SLINKY MODE?

Edit: When I think more about this, the "merge" is actually moving forward (TARANTULA) as originally I presumed... that's the only way it can possibly work. (I think I need a beer...)

 

Have to say, I always assumed it would do it that way - seems the simplest solution (perhaps only solution as you say).

 

I guess we'll have to wait for the documentation to be sure. If I'm using the SYNTHETIC FULL check box, the incrementals following the baseline start to disappear and when I look at the date of the baseline under Macrium/Restore window, it now has the date of the incremental that just disappeared. That makes me think things are moving forward (aka TARANTULA MODE). BUT if you look at the creation/modification dates under Explorer, they remain the same as when the baseline was first created.

I'm gonna watch the size of that baseline and its 1st following incremental, as this process moves forward, to see if I can spot anything in particular. I really need to know the status of that BASELINE.

 

That would actually make sense if you're not using the SYNTHETIC FULL mode... the baseline would remain and the incrementals themselves would crawl based on their retention number.

 

Pete

Great tip. In the past with v5, I put a txt file in the xml containing folder that described each backup. Now I don't have to do that.

Thanks for all of your input on this exciting new imaging program.

Silver

 

I don't have anything of value to add to the discussion, other than my appreciation of all the hard work you people are doing that is of great interest and value to those of us who don't have the knowledge to do it for themselves. I'm considering v6 for at least a secondary imaging program to Shadow Protect. As I follow the discussion here, that might change to v6 as the primary and SP as the secondary. Again, thank you.

 

AO, just jump in! The water's... uhhhh, frickin' cold!!

 

I dunno, seems pretty complicated to me.

 

I've been following this thread silently and testing v6 on my end. Very pleased with the result so far. I didn't stretch it as far as some did (a seemingly very flexible frog), but used it as my normal day would be.

 

What... you actually have a NORMAL DAY? (man, I wish I had one o' those)