Discussion in 'backup, imaging & disk mgmt' started by garry35, Jun 4, 2015.
is it safe to use rollback on SSD drives, by safe i mean does it shorten the lifespan of the drive
Yes. Any build from the last build of RollBack Version 9.1 onwards supports SSD drives.
If you have any questions regarding SSD Trim support I suggest reading our forum post which goes into a great amount of detail.
Hi Garry! I know Kurtis means well above, but until we hear more officially from Horizon DataSys on this issue, I suggest you read this recent THREAD concerning your very issue.
At the moment it has been shown that Rollback, currently, does not properly support SSDs. They are investigating this finding and hopefully will report back soon on this issue.
The question is regarding SSD's which RollBack Rx supports. That thread is in regards to SSD trim, which we're still investigating. However if this poster is curious about simply installing onto an SSD then see my above response.
Kurtis, that's not what the OP asked... the Op stated, "...by safe i mean does it shorten the lifespan of the drive."
Without a functioning TRIM command being received by the SSD, using Rollback will, indeed, shorten the life of a protected SSD.
I referred to the ongoing investigation and if the OP cares about his SSD and its overall life, I would suggest he wait for the results.
Depending on how you use it (number of snapshots, their sizes how often you take them, etc.), it will impact the wear-leveling algorithm of the ssd; some cells will deteriorate much faster than others.
Froggie, trim is not related much with the lifespan of the ssd, as long as there is plenty empty space for over provisioning.
SSD Trim is not going to be a major factor either way in determining the lifespan of the SSD. The SSD Trim issue is still being looked at by our development team. Regardless, I stand by that you can install no problem. Whether you want to trim is up for debate, but it's still possible to install and start running. If you're really worried about SSD Trim, disable it for now until we find out definitively.
Not entirely true:
- If the user has left 20-30% of of the disk as unallocated space, then yes Trim is not a factor.
- But, if the user has left zero unallocated space, then Trim is essential for the ssd lifespan.
Can you explain what do you intend with disable trim?
If you intend the trim of windows throught the "fsutil", there is no reason to do it because RollbackRX driver blocks the OS trim commands anyway.
Fact of the matter is that we have many users who use SSD's and are not reporting a problem. If one exists, it's a small group, and we'll correct it. However, I think if you are to go ahead and install RollBack Rx on a SSD you should be fine.
Funny thing is, I have an SSD and a traditional HDD at home I house my data on the regular HDD and house the OS and RollBack Rx on the SSD. I've honestly not had any issues regarding SSD Trim and RollBack Rx. So, if Froggie's suspicions and setup prove true, it's probably a combination of many things.
We'll just have to wait and see I suppose
It may appear as if it's not working, but it masks it in a sense. It can still function, but if I'm not mistaken it's just the appearance of it not working.
Gotta love the debate.. Fact of the matter is RBRX does not support the necessary functions for proper operation of an SSD. Not yet.
Depends on the setup. Everyone has a different setup, and your statement is too general to really cover all situations.
Many users use SSD's, and perhaps there is some small issue, but I would never go so far as to say that RollBack Rx won't work on all SSD's.
Really, if you don't want to take the risk then wait until we publish our findings, but I have many users who use the software and SSD trim with no issues.
Froggies demo proved that (at the very least) on his system TRIM was not working and HDS agreed this was the case. At a minimum this should send up warning signs and it is totally irresponsible (but not surprising) that a HDS employee is saying that caution is not warranted. There is no way that Kurtis can know this for a fact at this point. The fact that few people have reported issues simply demonstrates the possibility that those using SSDs have not noticed the problem yet, after all it is a problem that will take time to develop, it does not mean there is no problem.
The suggestion of waiting for further information regarding this issue is good advice.
I disagree, I do not think Kurtis means well at all, he is simply promoting sales of his employers product. Thats what he is paid to do. Given the track record of HDS in regard to issues with Rx and addressing (or rather not addressing) these issues it comes as no surprise that Kurtis has dismissed this issue out of hand.
can we finally have the truth about trim from HDS and verified by an independent source. i understand that HDS doesnt want to give away its secrets, but surely this could be covered by some non disclosure clause or agreement by the parties involved and give everybody peace of mind and maybe put back a little trust in HDS instead of doing the opposite
I did that already. How much more independent do you need?
I have thousands of customers who use SSD, Trim, and RollBack Rx with no problems.
Kurtis, enough with the nonsense!
RollbackRX drivers block any and all trim commands.
If you want to continue with your "marketing answers" do it in your own site and in your forums, not here!
ps. And if HDS wants to make us believe, that RollbackRX "trims" the ssd make your own live demonstration to prove it...
We're still investigating the issue Froggie presented, if we find anything that needs changing we will make the appropriate changes on our end. Until then, the official stance still stands.
Like I've said a few times now, issue here is with TRIM. If you still want to install RollBack Rx on an SSD then you can continue to do so.
I agree whole heartedly with Pandlouk about this. You know you can't back it up with actual facts so leave the marketing home.
Interesting exchange of opinions vs facts
Exactly what would that issue be. Because TRIM has no issues. RBRX in combination with TRIM has issues.
Well, I've clearly demonstrated to HDS Senior Technical people that SSD does not function on a Rollback enabled system... I can do no more.
Kurtis' understanding of SSds and their functioning lacks quite a bit of basis in fact... but he's a "marketing" type so I don't hold it against him.
If there are any Rollback protected SSD owners who are doubters and wish to find out for themselves (feel free to try it, Kurtis) whether TRIM is functional and whether lack of such affects SSDs, try the following suggestion (at your own risk, of course)...
Most SSDs whose over provisioning space is not programmable come with about 7% over provisioning space. Do the following...
Fill up your protected SSD with data (all partitions on the SSD, protected or not, all but a coupla gBs) then take a snapshot. Then delete all that data and take a snapshot. Hopefully Rollback will make that space available once again to Windows (look at your C: partition space). If it did, fill it up once again.
At this point, your SSD will be filled with the original fill-up data (unTRIMmed) and will now try and find space to add your new fill-up data... it will not find enough in the little space left (Windows unAllocated) plus whatever over provision space might be available. When it runs out, it has no choice but to start executing READ/MODIFY/WRITE operations on its NAND blocks, filling up those single bytes with data that needs to be written... this will take a long time, The R/M/W operation is very slow... and it will have to do millions of these causing excessive WRITE AMPLIFICATION of the NAND block being RMWed. This WRITE AMPLIFICATION will start decreasing your drives life at a fairly great rate.
The only solution at this point (and the drive may bind up on you during the process) is to unInstall Rollback (and you must reBOOT following this operation no matter what the system lets you do) and hopefully Windows (7 with the SSDtool or 8 with OPTIMIZE) will be able to get the drive back on track... hopefully. If not, it'll have to be SECURE ERASEd and you can start from scratch.
What people don't realize is that as you create snapshots and eliminate them, the Windows allocation will gets its "space" back but the SSD will not without TRIM. Eventually it will fill up and you'll experience the above phenomenon.
That's a s far as I'm going to go with this... I've successfully demonstrated this "event" to HDS Senior Technical People (and it's not a Frog system anomaly... I can do it on any SSD-based system) and hopefully they'll see the bug and fix it... hopefully.
I've done all I can at this point to bring this issue to HDS' attention... I've definitely got other things to do (Drive Cloner v6... I'm gonna break your back )
That is again correct.
In even more basic terms, an RBRX system doesn't fully complete garbage collection. And this strains the contemporary SDD far more than normal. RBRX does not let the SSD get anything back. The SDD is falsely led to believe that each protected sector is permanently in use.
Using TRIM as intended would be the equivalent of using CheckDisk or a Linux boot disc to modify the volume of RBRX. We all know that is a no-no.
As an exercise I went through some proof of concepts and it can definitely be done. Hopefully they will arrive at a solution similar to what I was playing around with.
Separate names with a comma.