Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Full Restore Speed
#1
Why is the Hasleo Backup Suite full restore speed faster than Macrium Reflect X even in Hasleo Backup Suite free version?  I upgraded my Macrium Reflect v8.1 perpetual license to Macrium Reflect X annual subscription but their full restore speed is still twice as long as the full backup speed.
Reply
#2
(10-09-2024, 12:37 PM)Keith Weisshar Wrote: Why is the Hasleo Backup Suite full restore speed faster than Macrium Reflect X even in Hasleo Backup Suite free version?  I upgraded my Macrium Reflect v8.1 perpetual license to Macrium Reflect X annual subscription but their full restore speed is still twice as long as the full backup speed.

This question is difficult for us to answer, we haven't looked into how Macrium implements the full restore feature, all we can say is that the speed of full restore has a lot to do with the algorithm.

Have you compared the performance of Macrium Reflect X and Hasleo in backup and Delta restore?

Thanks.
Reply
#3
(10-09-2024, 12:37 PM)Keith Weisshar Wrote: Why is the Hasleo Backup Suite full restore speed faster than Macrium Reflect X even in Hasleo Backup Suite free version?  I upgraded my Macrium Reflect v8.1 perpetual license to Macrium Reflect X annual subscription but their full restore speed is still twice as long as the full backup speed.

Aren't you happy ? Big Grin
Reply
#4
In Macrium Reflect X I need to add LegacyRestore DWORD 0 to HKEY_LOCAL_MACHINE\SOFTWARE\Macrium\reflect\Settings every time I boot from the Rescue Media to get the full multithread restore speed during non-delta restore otherwise it takes twice as long to restore than it takes to backup. Macrium Reflect X disables multithread restore by default if either source or target is a USB device due to problems with many USB 3.2 Enclosures or Hubs not coping with high-bandwidth sustained read/writes and causing abrupt disconnections in the middle of a restore or clone.
Reply
#5
Keith, you may find the exact same problem that Macrium REFLECT occasionally experiences, with enclosures and hubs.  It is a bandwidth problem with the devices and has nothing to do with the application... other that the application using as many cores/threads as it can to do its job.

Both Macrium and Hasleo use that approach for speed... expect the unexpected.
Reply
#6
Yes, it is important to use multithreading, which can significantly improve performance. Has anyone compared the backup performance of Macrium Reflect X with Hasleo?
Reply
#7
Reflect X vs Hasleo v4.9.3


CPU: i7-10700, 8-core/16-thread, 2.9/4.8ghz 
RAM: 32gB
Source disk: NvME M2.2280, PCiE v3 (7-part <1 [email protected], 1 unf@16mB, 1 [email protected], 1 [email protected],
Target disk:  NvME M2.2280, PCiE v3
MEDIUM compression, no image checks following imaging operation

Hasleo v4.9.3: 4m 57s (Image size=83.5gB
Reflect v10: 3m 9s (Image size=71.4)

Remember, Macrium REFLECT X is now using an entirely new streaming compression method, when imaging, obtained from a Facebook developer.  As a result, the image format (not the method) is now PUBLIC and available to all if interested.
Reply
#8
Reflect X vs Hasleo v4.9.3


CPU: i7-10700, 8-core/16-thread, 2.9/4.8ghz 
RAM: 32gB
Source disk: NvME M2.2280, PCiE v3 (7-part <1 [email protected], 1 unf@16mB, 1 [email protected], 1 [email protected],
-------------------------------------------------------- 1 [email protected], 1 [email protected] & 1 [email protected])
Target disk:  NvME M2.2280, PCiE v3
MEDIUM compression, no image checks following imaging operation

Hasleo v4.9.3: 4m 57s (Image size=83.5gB)
Reflect v10: 3m 9s (Image size=71.4)

Remember, Macrium REFLECT X is now using an entirely new streaming compression method, when imaging, obtained from a Facebook developer.  As a result, the image format (not the method) is now PUBLIC and available to all if interested.

I'm not sure if Hasleo is able to determine the USED SECTOR amount with EXT<n> images... that may be the main reason for the differences in image sizes.  I'm sure the ADMIN (Dev Team rep) will jump in here and correct this assumption, if necessary.
Reply
#9
(10-11-2024, 12:33 AM)Froggie Wrote: Reflect X vs Hasleo v4.9.3


CPU: i7-10700, 8-core/16-thread, 2.9/4.8ghz 
RAM: 32gB
Source disk: NvME M2.2280, PCiE v3 (7-part <1 [email protected], 1 unf@16mB, 1 [email protected], 1 [email protected],
-------------------------------------------------------- 1 [email protected], 1 [email protected] & 1 [email protected])
Target disk:  NvME M2.2280, PCiE v3
MEDIUM compression, no image checks following imaging operation

Hasleo v4.9.3: 4m 57s (Image size=83.5gB)
Reflect v10: 3m 9s (Image size=71.4)

Remember, Macrium REFLECT X is now using an entirely new streaming compression method, when imaging, obtained from a Facebook developer.  As a result, the image format (not the method) is now PUBLIC and available to all if interested.

I'm not sure if Hasleo is able to determine the USED SECTOR amount with EXT<n> images... that may be the main reason for the differences in image sizes.  I'm sure the ADMIN (Dev Team rep) will jump in here and correct this assumption, if necessary.

Yes, HBS doesn't support EXT partitions, have you tested not backing up EXT partitions? Sorry we were held back by some issues with other products and didn't have time to test.
Reply
#10
(10-11-2024, 11:28 AM)admin Wrote:
(10-11-2024, 12:33 AM)Froggie Wrote: Reflect X vs Hasleo v4.9.3


CPU: i7-10700, 8-core/16-thread, 2.9/4.8ghz 
RAM: 32gB
Source disk: NvME M2.2280, PCiE v3 (7-part <1 [email protected], 1 unf@16mB, 1 [email protected], 1 [email protected],
-------------------------------------------------------- 1 [email protected], 1 [email protected] & 1 [email protected])
Target disk:  NvME M2.2280, PCiE v3
MEDIUM compression, no image checks following imaging operation

Hasleo v4.9.3: 4m 57s (Image size=83.5gB)
Reflect v10: 3m 9s (Image size=71.4)

Remember, Macrium REFLECT X is now using an entirely new streaming compression method, when imaging, obtained from a Facebook developer.  As a result, the image format (not the method) is now PUBLIC and available to all if interested.

I'm not sure if Hasleo is able to determine the USED SECTOR amount with EXT<n> images... that may be the main reason for the differences in image sizes.  I'm sure the ADMIN (Dev Team rep) will jump in here and correct this assumption, if necessary.

Yes, HBS doesn't support EXT partitions, have you tested not backing up EXT partitions? Sorry we were held back by some issues with other products and didn't have time to test.
Rerunning the above test with the Linux (EXT<n>) partition out of the mix, the results are as follows...

Hasleo v4.9.3: 3m 41s (Image size=71.87gB)

Reflect v10: 2m 04s (Image size=64.13)
Reply


Forum Jump:


Users browsing this thread: 7 Guest(s)