differential backup with automatic cleanup?
I'm trying to do a differential backup at regular intervals, making a new full every 5 diffs, storing no more than 2 recent version chains. However it is now failing to 'consolidate' at a time when it shouldn't be trying to do that (only 4 diffs of the latest version b3; it should be trying to write b3_s6_v1, if I have any understanding of what's going on, and not be trying to 'consolidate' anything.
It appears to have successfully removed b1; the consolidation failure is allegedly because it can't get exclusive access to a file whose name belongs to a completely different backup on another disk.
Is this collection of functions known to work properly in general? The behavior is as if the config files had scrambled names in a rather random manner.
Avery,
Please start again with new backup tasks.
1. Be absolutely certain that the backup scheme reads Custom/Differential and not Differential/Differential, or
Custom/Incremental and not Incremental/Incremental
2. Assign a name to the backup task which has not used before. Task name to be unique.
3. Point the task to a new folder or sub-folder. Use a folder which does not have other task feeding.
----------------
If configured to full plus 5 dif or 5 inc and store 2 recent versions
You can expect the first deletion to occur afer Full B3 and delete after each new FULL in the future.
My recommendation is to avoid editing an existing task as rarely ever does the edit produce the results you expect. Start fresh with new tasks when corrections needed.
Curious. Which version of Windows? You have not mentioned that in any of your postings I have seen.
Thanks!
Redo all of them? That seems excessive. Starting a new one following 1-3 does not fix the problem! For a while, with the earlier backup, the backup operations seemed to be working in spite of the error messages, but then they began failing due to 'failure to read from disk', even though both the to and from disks were available to windows explorer.
My OS is Win 7 Ultimate.
I'm definitely inclining towards rolling back to ATI 2014; ATI 2015 is really looking like something that got shoved out the door prematurely because a manager said it had to be.
My suggestion was and is to not use the tasks which are not woriking correctly. You only need new tasks if you want those specific tasks to work.
If I were getting disk read errors, I would check every partition (visible or hidden) for disk errors and would check each disk involved with the backup.
Assiign a drive letter temporarily to any hidden partition so that partition can also be checked for errors.
If you decide to go back to 2014, I would suggest that you run the TI cleanup 2015 or 2014 cleanup utility (signaure link 9 below) and perform the text instuctions about checking he registry. New tasks will be needed if you revert to 2014 as the cleanup utility will delete the old tasks.
Continuing to fiddle, I think part of the problem is that backing up the USB (3) drive fails if any removable drive that was present when the backup was created is removed, even tho these drives were not included in the backup task as specified in the interface. Making new backups to the same disk & partition did not eliminate this effect, but it stopped happening when I made a new backup to a different disk and partition, and now I've managed to get it apparently working for the one I wanted. So the program seems to sometimes retain and use some information it shouldn't be looking at in inappropriate ways, but eventually forgot it, in this case.
On a related but different topic, it strikes me as odd that for differential backups, there isn't any option to just keep the last n differences from the original, which is retained unaltered. Logically, it seems to me that it should be possible to remove by hand earlier diff files without messing things up, but I've seen enough odd behavior so that I'm not sure that it's true.
When you remove the non-related drives causing the failure, what is the drive letters of target disk and what is the drive letters of the non-releated disk removed?
As to deleting diff, I have had success in tests where I manually deleted Dif-S2 and Dif-S3 and keeping Dif-S4. The key is to simualte a restore or validation immediately following the deletion so the program will prompt you to respond to its promps to ignore the missing. If two missing, you will need to respond twice to its prompt to ignore.
Some issues with usb sticks or drives can be the timeout settings in the Device Manager/Universal Serial Bus Contoller/
Either the Generick USB Hub, and/or USB Root Hub
On all hubs listings, click the Power Management options
Uncheck the "All the computer to turn off device to save power."
Also, there is the Contro Panel/Power Management which has settings to run off usb devices, these Windows Power Management settings must be disabled.
The backup is of a USB 3 drive lettered I:
To a backup folder in a partition lettered G: (on a platter HDD), where the problem first arose, or
To a backup folder in a partition lettered C: (on an SSD)
The drives that ought to be irrelevant are a USB 3 drive lettered J:, and an esata drive lettered J:
So the problem will arise when only the drive being backed up is USB, the one that ought to be irrelevant and the place where the backup is stored both being sata (the motivation for doing this is to take frequent backups of the removable drive which I carry around a lot & almost lost a few days ago).
It is not a good feature for a differential backup to care about anything other than itself and the first version it is diffing against, and furthermore for a supposedly simple interface to require sophisticated knowledge of the program to deal with missing earlier diffs, which can be produced by automatic cleanup if you select the 'limit backup size' option.
I'm not surprised that compex setting changes such as between differential and incremental would cause trouble; changes in scheduling and maximum backup size don't seem to have caused problems for me.
Avery,,
I'm another user like you. The program allows for editing a task (but with no instructions or limitations) and 2015 provides no options to delete a single tib files. My comments are just from my experience as to how to make it work for me. What works for me may not work for you. I most certainly agree that any single diff file should be recoverable or validate if the original full is available--both without errors. In past versions, if any diff file was missing, the file would not validate. It has only been within versions 2014 and 2015 that a diff file will still validate even with other diff files missing.
If you wish Acronis to know about your views, you need to notify them as there is no guarantee that Acronis sees your comments when posted i n this commnunity forum.
Thanks! (but if they can't be bothered to keep an eye on their own user forums, I'm somewhat inclined to just sit back and let some competitor such as perhaps HDClone overtake them ... Acronis really did save my butt once, but gratitude only lasts so long ...). Anyway, my scheme for semi-continuous backup of a flash drive seems to work reasonably stably, & your comments have been very helpful for getting it to that state.
One final thought on this issue. When a task is created, the program ties the both the source and Disk ID and drive letters to its requirements. Thus, if the insertion or removal of a usb stick prior to a backup causes the target disk to have a different drive letter, the task will fail.
One test you might want to try would be to
Assign a specific drive letter ( such as x or y or z) to a freshly formatted flash drive.
Create a new scheduled task pointing this to that special lettered flash drive.
Add or remove another flash drive before the backup.
See if the new task pointed to the special drive letter runs on schedule.
Note: the drive letter assignment has to be done on the computer which has TI installed.