Automatic cleanup doesn't work

There have been some issues with the auto-cleaning options of 2011. From experience:
- do not edit the rules after you have set them for a fresh task, create rather a new task,
- the option that works the best is to store the X most recent version chains.

You have enough space for all the backups you want to keep plus one one more? You need the space for one more so that ATI can make the backup before running the limits rules and deleting any backups -- it's done that way so you don't accidentally delte all your backups if soemthing goes wrong with the backup process.

Scott,

If what you say is correct, this is a pretty severe depreciation of the software from previous versions, and I would not expect it to work that way.

Pat,

I have created entirely new backup processes multiple times. My latest attempt to resolve this on my own was to stop using incremental backups, and go with full backups each night.

ATI has always worked this way; you always needed space for all the backups you want to store plus one more.

Suppose something is preventing the backup from completing properly -- if ATI did the rule check before the backup, it would have already deleted a backup -- run multiple times and one could then actually end up deleting all of one's backup with getting any new backups. To avoid these undesireable outcomes, ATI does the rule check after the backup is completed. The price one pays is needing room for one more backup than the rule limit.

That's simply not true. I have a machine running ATI Home 11 which behaves the way I've described and works flawlessly for months on end. It occasionally leaves a stale backup behind from several days previous that has to be manually deleted, but all in all it's reliable.

James,

Did you try:
- using custom incremental,
- do a new full every x incrementals,
- auto-cleaning on,
- keep only y most recent backup chains.

When estimating space, building on what Scott's indicated, remember that ATI will do a new full before erasing an older chain. In other words, if you x to 6 (typical for daily backups) and y to 1, you will get:
- backup.tib (full)
- backup2.tib ... backup7.tib (6 incrementals)
- backup(1).tib (new full), *then* ATI will delete backup.tib and all the incrementals based on it. Then, next time,
- backup(1)2, etc.

Rule of thumb is that at a minimum (case above) you would always need backup up storage space equals to 2.5 to 3 times full backup. This rule of thumb works pretty well regardless of the scheme you use.

Yes, I tried all of those options except the last (keep only y most recent backup chains), which I am avoiding because it's not automatically scalable, and this needs to be a hands-off solution. When both users had ATI 11, everything worked wonderfully. User1's computer was replaced with a new Windows 7 machine, so they bought ATI 2011.

I last configured the machine in question for full backups with a 90GB limit on the backup location's usage.

Here is the backup server (essentially just NAS):

server:/media/raid# du -h
94G ./User1 (using ATI 2011 update 2, build 6868)
16K ./lost+found
73G ./User2 (using ATI 11)
166G .
server:/media/raid# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 458G 98G 337G 23% /
tmpfs 248M 0 248M 0% /lib/init/rw
udev 244M 124K 244M 1% /dev
tmpfs 248M 0 248M 0% /dev/shm
/dev/mapper/nvidia_fcfffejj1
184G 166G 8.3G 96% /media/raid
cstech-srv:/media/raid#

Here is User1's current configuration is attached to this post. Since it is set for a 90GB limit with automatic cleanup, why is 94GB in use at the backup location? This is as of this evening, following a successful backup of User1's computer, and a failure on User2's because there wasn't enough free space to perform a backup.

Edit: I think I see now that you are correct about ATI deleting the oldest backup after creating the latest, the reason they never had a problem was that one would backup long before the other, and the backups were smaller, so the available space on the share was plenty for the temporary storage of the latest before the deletion of the oldest. Is there any way at all to change this behavior? If I did compromise and limit the number of backups instead of the size, would it delete the oldest first? Is there _any_ scenario in which ATI deletes the oldest backup first? Storage comes at a premium, as you can see above. User1's full backups are 32GB each, so we're pushing it already.

You could use Windows task scheduler to create a task that:
- deletes everything in the backup directory, then
- launches ATI task.

Alternatively, you could add a pre-command to the backup task to first blank the directory, then do the backup.

If you have the time, investigate using Chain2Gen which is a helper program for TrueImage Home. Detailed information about C2G can be found by clicking on my signature link below and looking at all the times in index item #5.

The purpose of C2G was to provide the user with more controls over how many backup chains are retained. C2G does not increase or decrease the number of tasks that you might use.

Understanding what C2G does and does not do is key to understanding how it works.

All scheduling of backups and creation of backups is performed as a result of the TrueImage backup task. The function that C2G supplies is to run as the beginning part of the TI Pre/Post Pre function. Its only real purpose is to ask and answer this question:
"Is now the time to "empty" the contents of the backup folder so a new full backup will be the next creation." If the answer is no, C2G closes with no further action and the TI begins the backup. If Yes, is the answer, C2G will empty the set0 folder. The term "empty" means to rotate the contents into history. All TI backups are made into the set0 folder with C2G controlling when the contents of the set0 folder is rotated into history. The user sets the C2G parameter file as to when the set0 folder is emptied via rotation upward. The amount of history retained is user controlled by parameters set in the parameter file. History is the prior *.tib storage folders named set0,set1,set2,set3, etc. based on how much or how little history is to be retained.

The easiest implementation is to follow the my guide listed in signature index item 5-B. You will need to change the illustrated parameter to match your needs. Remember, the function of C2G is to help you control how many chains you wish to retain.

Some extracts from the parameter file as to choices available to the user as to when the set0 foler is emptied or rotated into history.

:: ---Examples begin here----------------------------------------------------
:: option 1 set a trigger based on total number of .tib files in folder
:: 0 means unlimited incr
:: 1 means a new full every time (1 Full + 0 incr = 1)
:: 2 means a new full after 1 incr is done (1 Full + 1 incr = 2)
:: 3 means a new full after 2 incr is done (1 Full + 2 incr = 3)
:: The thought behind this control is the exposure risk of an incr chain is a
:: function of the chain length. Hence start a new chain when the length
:: of the chain exceeds some threshold. (or set 0 for manual control)
:: TRGR_ON_MAX_TIB_CNT=[0-N] 0=unlimited N=max_number of ".tib" files in chain
:: --User must set this next variable if the preceding option 1 is applicable--------------
set TRGR_ON_MAX_TIB_CNT=0
:: ---------------------------------------------------------------------------

:: ----Examples begin here----------------------------------------------------------
:: Option 6 set a trigger based the SUM of the INCR file sizes exceeding a users threshold.
:: The thought behind this control is regardless of how many gig the full backup
:: might be, it is the total MB of the incrs which represent the exposure risk.
:: Hence start a new chain when the "weight" in incrs exceed a MB threshold.
:: We will do this regardless of how many incr files it takes to exceed the threshold.
:: TRGR_ON_TOTAL_SUM_INCR_MB=[0-N] 0=unlimited N=max_sumsize in MB
:: ----User must set this next variable if the preceding option 6 is applicable--------------
set TRGR_ON_TOTAL_SUM_INCR_MB=0
:: ---------------------------------------------------------------------------

I too had this problem and this is why I went away from ATI 2011 and then ran into the problem with the uninstall and found this problem ( http://forum.acronis.com/forum/14782#comment-68844 ). I recently downloaded the latest version and tried the backup process and to this point the auto cleanup is working fine so I think I will finally stay with ATI. Although I like MS backup option, but it doesn't seem to have all the features as ATI.

That said, I don't know if I were to uninstall if I would run into the problem like reported in the link above.

-Grant

I have had the same problem as James. Running 2011 TI on 2GB external drive for backups and running Vista 64. Have set version control scheme. Have found the "delete last backup function when not enough space left does not work on backup drive. I only use this backup drive for backups. I do not have enough backup literacy to understand the recommended fixes in the previous posted comment chain. Is there a way I can go directly into my backup drive and manually delete all data on this drive not associated with specific backups I want to keep. I am running version chain schema so don't want to delete any data off of drive that will screw up version chain of backups I want to save.....Short version - what specifically can I delete and not delete off of my backup drive - how do I tell from folders/files on backup drive all data associated with validated backups I want to keep or delete that are shown in 2011 TI previous backup list.

Shamuslad,

PAT L in post #1. wrote:

There have been some issues with the auto-cleaning options of 2011. From experience:
- do not edit the rules after you have set them for a fresh task, create rather a new task,
- the option that works the best is to store the X most recent version chains.

This advice continues to be the best advice.

As Pat stated,your best chance of success is to use the custom scheme and keep x number of chains. The last 7046 build of 2011 fixed many of the cleanup issues but keeping x number works much better than attempting to use a space or time rule to control the number of backups. A new task directed to a new storage folder works best.

Here are 3 examples. Adjust the number of backups to what you want to keep but any of these three has been tested to work very well.

Remember the characteristics of the incremental backups. If any of the older incremental backups become corrupted or non-usable, then all the newer incremental backps are useless. It is better to have more frequent fulls rather than a large number of incrementals. Whereas, when using full +diff,you only need the one full plus an optional single diff to be able to restore.

Full backup scheme

First Full plus subsequent Incremental backup scheme

First full plus subsequent Differential backup scheme

Login and download the latest version/build. They resolved a lot of problems.

Grant, thx for suggestion. Already have update 2011 to latest version and it did not resolve the issue with deleting the oldest backup when external backup drive does not have enough space. Do not know if this is because previous backups were made with older version of 2011 TI.

IF ati deleted a backup before it made the next backup, and there was some problem preventing backups from being created properly,-- a disk bug or whatever -- you could end up deleting all your backups as you retired to backup and have not valid backups to show for your troubles.

In any event, destroying data before replacment data is in place is generally a less conservative protocol and one that ATI has not followed in ati going all the way back to ATI10, the first version with the auto-filemanagement features.

I would delete the backup schedule and start a new one. Keep the backups until it is safe to delete them, but see if your problem is resolved. I started with only 2 incrementals and 2 backups sets and then required it to delete. It started working fine so I increased the incrementals to 10. No problems since. Hope you are able to resolve.

Thanks again, Grant. I went to original backup that was only for data drive (not system drive) and deleted it. Also reset backup scheme to full backups (without incremental/differential) for entire PC disk (data & systems - same as newer previous backups) and set flag to delete last backup when not enough space. Doing this provided me enough drive space to do new full backup I was able to validate. After I see if oldest full backup is deleted when storage space becomes an issue, I will change scheme to using incrementals like you did. Start with low number and increase it.

Thx again for you help.....