Skip to main content

TI 2020 functionally useless - Far too slow

Thread needs solution
Beginner
Posts: 1
Comments: 0

Sorry guys but all backups take too long now. I backup through a usb3 hub . I've bought a new disk, replaced the usb 3 hub. I've started all my backups from scratch (deleted all the existing backups) It's the same story everytime. I just want to get back to how it was. I'm runing on Windows 10 64bit.. upto date Maybe it's a Windows 10 issue? Please advise if best option is to ditch current release or is some fix about to be released?

1 Users found this helpful
Forum Hero
Posts: 36
Comments: 7580

#1

Mick,

You're not alone in experiencing slow backups.  Are you including validation in your backup task?  Are you creating incremental backup schemes?

Answers to these questions will help.

Beginner
Posts: 0
Comments: 1

#2

I have the same problem. An incremental, a day after the full backup, lasts more than an hour, and I had to stop it because it takes too long.

I don't trust the backups of TI2020 anymore.

 

Forum Hero
Posts: 36
Comments: 7580

#3

To little information to comment anything meaningful.  More details on task configuration, number of backups, typoe of computer to begin with is necessary.

Beginner
Posts: 3
Comments: 5

#4

Sorry, Enchantec, but I have this problem too and I think we can all agree that we are comparing like with like - TI2020 v T2019 is incredibly slow and our configurations etc are unchanged so your comment is irrelevant. Could it be a deliberate Acronis policy to wean us off tib to tibx? I haven't tried yet to see if the latter is actually faster.

Forum Star
Posts: 147
Comments: 966

#5

Actually, Enchantech's questions are not irrelevant if you want suggestions on how to speed things up (such as not doing validations.  Answers to those questions also clarify the problematic context which can be beneficial when reporting the problem to Acronis support.

Forum Hero
Posts: 36
Comments: 7580

#6

I cannot comment on the workings of TI 2020 with tib backups as I have none.  So I cannot comment on the performance of the app when creating files of tib format.

I opted to make a complete change to the new tibx format when TI 2020 was officially released.  I retained what current backups I had from TI 2019 and if I had to restore any of them (not even remotely anticipated) I would use the TI 2019 Recovery Media to do so in the case of a full disk restore.  For a file or folder I might use either the recovery media or the installed TI 2020.

I have found performance of TI 2020 at least on par with TI 2019 and in several areas superior to TI 2019.  Why are my results different than others?  Have no idea really.  Could be any number of things.  Having said that I do have some suspicions.  One is that most users of True Image these days are doing so on laptop machines.  Most of these are general consumer grade machines from the likes of Dell, HP, Acer, MSI, etc.  In my experience with these devices I find them of inferior quality and the majority are using less than performance grade hardware.  Couple that with the fact that a good number of users machines are a few years old adding to the performance equation and I suspect that some users simply have machines that do not work well with TI 2020.

I am not making excuses for Acronis having said the above.  There may well be other factors involved in many cases.  I only speak of my own experiences.  I do have three laptops, all of them are several years old.  Only one of those machines has TI installed on it.  I only use it when I travel so it does not see a lot of use.  It does not have a lot of software installed on it either.  This laptop is a Lenovo Flex 15, has an i7 CPU, upgraded DDR3 dual channel ram (16GB), and runs two SATA SSD's one for the OS and one for data.  TI 2020 works just fine on it and backup is comparable with TI 2019.  When I do backup this device I use a non-scheduled, Custom scheme, Full method setup.  I backup the OS drive to the internal data SSD and then later I move that backup to an external drive.  I use Recovery Media to backup the data drive direct to the external drive.

All other installations I have of TI are on desktop machines that I custom build myself.  For the most part these are very high performance enthusiast machines running Gen 3 NVMe M.2 drives in multiples.  I do have one older desktop that is now some 11 years old that I do not run TI on.  It belongs to my spouse and I can rebuild it easily from scratch so I see no need to even back it up.  My other machines run TI fine and performance is where I expect it to be.  Having said this I will add that TI 2020 behavior is different than that of TI 2019.  For example, once you click the Backup now button the app spends a good deal of time scanning and calculating data.  This has to do with the meta-data gathered for the data being backed up and the creation of the programs internal database for the backup which tracks the data backed up as well as the backup file itself.  The reverse of this can be witnessed during recovery as well.  To the user this seems like the app is hung or not responding and on a slower machine probably takes more time than users are accustomed to.  Another area where I have found a true lag in performance is in networking although I am not convinced that it is True Image at fault here.  I run a complex home network in which I have two networks working together to which all of my high performance machines are attached.  When I initially setup connections in TI to other devices on this network it is necessary to wait on the app to resolve the device location before authentication occurs.  Once it has done that however all work fine thereafter.

I am confident that my answers here mean nothing to you or your situation and I fully understand that.  Is there something inherently wrong with TI 2020 that hobbles performance?  In my opinion no.  Obviously in your opinion it does.  I suspect we can agree to disagree on that subject.

Forum Member
Posts: 7
Comments: 32

#7

Hi Enchantech, I have a Lenovo P72, i7 8750H, 16GB DRAM, 3 SSD: 1 Samsung M.2 NVMe 250GB, 1 Samsung EVO 870 M.2 NVMe 1TB and 1 Samsung EVO 860 SATA 1TB with ATI 2020.  It takes about 45min+ to complete backup of a 200GB partition from one internal M.2 to the SATA.  The validation also takes quite a while.  I really expected it to work much faster considering it's all internal and all.  Is that considered a reasonable time frame?  I was hoping for at least 20min.

Kind regards,

Michael

Forum Hero
Posts: 36
Comments: 7580

#8

Well, given the specs you have provided and using 45 minutes and 200GB as a basis it works out to a transfer rate of 75.95 megabytes per second using a 1024K denominator.  You are limited here to the SATA interface as well as the compression of the data which depending on the data type can have a lot of impact in certain cases, and the mean or average size of the total of all data.  This 75.95 MBps second equates to 607.6 megabits per second (Mbps).  Does that look the same as your Activity tab in the TI app?

For a Full OS disk backup I would be happy with that transfer rate given your machine specs.  There are a huge number of relatively small size files in an OS install and such files take much longer to compress and backup than do larger files.  That sounds counter intuitive I know but that is how it works.

Have you tried backing up from M.2 to M.2 drive?  If you can and do you should be in for a big surprise.  In an attempt to give you an apples to apples comparison I ran the following backups on my 5 year old Z97 ASUS Deluxe board based PC.  The rig has 32GB of DDR3 2400 RAM, an i5 4690 @ 3.5Ghz CPU (no overclock), with storage of:

  1. Samsung NVMe 970 Pro M.2 512GB for OS
  2. Samsung NVMe 970 Pro M.2 512GB for a scratch disk
  3. Samsung 850 EVO SATA SSD 500GB for app and user data
  4. WD Enterprise 2TB SATA HDD for Mass storage
  5. WD Black 1TB SATA II x 2 Drives one for archive, one for VSS copies

So this test involved a backup of the M.2 OS drive, all partitions Full disk, total data 79.1GB.  I first ran the test to the second M.2 scratch disk drive.  My results were a speed of 1,034Mbps or 129.25MBps with a total elapsed time of 5 min, 21 sec..

The second test I ran this same OS full disk to the Samsung 850 EVO SATA SSD.  My results were, are you ready for it, 1,140.3Mbps or 142.54MBps with a total elapsed time of 4 min. 51 sec..

Why is the backup to the SATA drive faster than the one to the M.2 drive I'm sure you are asking?  The answer is simple.  The Z97 ASUS board was the Gen 1 of M.2 design and featured a single M.2 slot that leveraged SATA Express channels to the PCI interface.  Thus the result reflects the speed of the SATA Express interface.

Your next question probably is, why are my results when backing up to an SATA SSD better than yours?  My belief is that first, I have less data on the disk.  The OS occupies around 50 percent of the total.  The rest is a few apps including TI of course.  Most apps and just about all user data are on the 850 EVO drive. 

For comparison, I have another machine which is based on an ASRock Z170 board.  This is a Gen 2 M.2 board having 3 M.2 slots on the mobo.  I have in two of these slots installed Samsung M.2 NVMe 970 Pro drives in a raid 0 for the OS drive.  I have two additional Samsung 970 Pro NVMe drives installed on add in M.2 PCIe cards in a second raid 0 volume.  All of these drives are 512GB drives.  The OS total data on disk is 73.3 GB.  A full disk backup of this array to the second raid volume ran at a speed of 1038.5Mbps or 129.8MBps with total elapsed time of 3 min. 29 secs.  This reflects the efficiency achieved by the Gen 2 board however, this speed is far below what the drives are capable of.  NVMe speed is measured in Gbps and the speed here is 1.0385 Gbps.  The PCIe Gen 3 x 4 interface is rated 3.2Gbps.  This shows how the small file sizes of an OS System drive drag on performance.

I hope this gives you some perspective on performance.  These results are real world and not benchmark results.  Another factor in these results is simply that even though both systems are a few years old they are based on high performance enthusiast boards and are factory optimized for performance.  The end result is better hardware equals better results.

 

 

Regular Poster
Posts: 60
Comments: 424

#9

Hello All,

Thought I'd add my speed comparisons for all to see and compare.

See attached Screen Shot of Full backup, differentials, Disk restore and One "oops" where I didn't have the Seagate Barracuda HDD inserted in the internally mounted hot swap bay when I started the backup.  All backups done manually.  NO Verification on any backup.

Note the Fast 1st backup on bottom, note the differential speed and note the last FULL backup speed near top.

The differential backup speed is really "pokie" ... Same hardware, same operator...

Use Ctrl with PLUS or MINUS to Zoom in on Picture.

Regards,

Steve F.

Attachment Size
527753-178815.JPG 63.99 KB
Forum Hero
Posts: 36
Comments: 7580

#10

Steve F.

Looked at your screenshot.  You state the the differential backup speeds are pokie.  I note that your differential backup files a small in size ranging from a few hunder MB to a few GB.  Note that speed is relative to total data backup size.  The smaller the data the slower the transfer speed.  This is to be expected.  Note as well the comparison of speed between the Full backup files.  There is notable speed differences between them and I assume that is because of a consolidation that took place of the previously created differentials to create a new full.  So roughly 10 minutes for consolidation and 25 actual backup time.  This all looks proper to me.

Forum Hero
Posts: 36
Comments: 7580

#11

For another comparison here I got out my laptop and checked on the performance of the last full disk backup I ran.  Again, this is between two Samsung EVO SSD's.  Machine specs are:

i7 4510U @ 2.0Ghz CPU

16GB GSkill F3-1600C9D DDR3 SODIMM

Samsung 850 EVO 250GB x 2 SSD's

Backup OS drive to data drive.  OS drive Full backup total data 46.4GB @ 503.6Mbps or 62.95MBps with total elapsed time of 6 min. 7 secs.

If this 6 min. 7 secs. is expanded out to 200GB the total elapsed time becomes 1 hour, 16 min. 9 secs. roughly. Edit: 52 min 55 secs.

Given this machine is a consumer level Lenovo laptop you can see that performance is not as robust as a desktop.  I attribute that to the hardware and design limitations of mobile devices.  I know that the industry and manufacturers want to tout their offerings as desktop replacements but that is all marketing hype.

Regular Poster
Posts: 60
Comments: 424

#12

Enchant,

Thanks for the insight !   I was not aware that the slower differential backup speed was a function of the quantity of data being backed up.  Obviously, what you are saying is true and  my screen shot shows it.

Is that actually by Acronis design ?  After ATI scans to log what files are to be backed up, one would assume that Acronis would use all the hardware horsepower available to accomplish the backup task in the shortest time possible.  It is almost as if they are "throttling" the backup speed intentionally ??

 

Steve F.

Forum Hero
Posts: 36
Comments: 7580

#13

Steve F.

I am not sure if this is intentional or if it is Windows functioning.  The use of VSS which has been around for ages in Windows works by generating a snapshot of selected data to be backup then, once the snapshot is created a VSS writer creates the backup or restore point based on that snapshot.  This arrangement was decided upon by MS engineers as a way to allow users to start a behind the scenes task and continue to work/use their PC while the task runs to completion.

So given the fact that TI uses VSS I would assume that any throttling would be a function of Windows.  Even if Acronis snapshot is used TI uses the Windows VSS system writer.  Could be enhanced somehow I suppose but I would think doing that would required TI to have its own dedicated VSS writer which I am confident they do not.

All in all, I really cannot complain about the performance of TI.  If you have a look into backup app reviews for the fastest you will find that I am not the only one who says Acronis True Image is the fastest.