IFW Imaging Speed

User discussion and information resource forum for Image products.
TeraByte Support(PP)
Posts: 1646
Joined: Fri Aug 12, 2011 12:51 am

Re: IFW Imaging Speed

Post by TeraByte Support(PP) »

Standard compression is not the best to test with when trying to check drive throughput. A better test would be using no compression (None) or Enhanced Speed - A (usually results in a backup image only slightly larger than Standard and is much faster).

I generally use Enhanced Speed - A for everything.
Kempy
Posts: 16
Joined: Sat Mar 01, 2014 9:29 am

Re: IFW Imaging Speed

Post by Kempy »

Thanks Paul,

I think we didn't want to check drive's throughput, we just wondering why ifw with the default/standard compression setting is 50% slower in imaging than the competitors.

As you suggested I found that 'Enhanced Speed - A' is realy a runner which can compete with the market in speed whether the compression ratio can't.

Images are 8-10% larger as with standard compression and compared to competitors.

Looking at the available compression settings in ifw shows, that the most are focused on compression ratio but only two on speed.

So a realy fast algorithm with good compression that can compete with the market is currently missing in ifw (?).

G.
TeraByte Support
Posts: 3629
Joined: Thu May 05, 2011 10:37 pm

Re: IFW Imaging Speed

Post by TeraByte Support »

You're not comparing apples to apples. The amount of data backed up, the
type of compression used, the strength of the hash algorithms. If you want
to test, enable VSS with all options checked, use Enhanced Speed A.

While VSS and enabling all the associated options (excluding files, etc..)
will be the fastest and smallest, you'd at least want to disable throttling
in phylock which gives priority to things running on the system that need
disk access.

Not only that, you can also control various items for the drives if your
drivers requirement. You can control the caching via the /po option (you
may want to enable the cache for those faster removable drives to start
with) and size of data read or written via the /iocf option. There is also
the /iobs option which gives finer control over it.


"Kempy" wrote in message news:9061@public.image...

Thanks Paul,

I think we didn't want to check drive's throughput, we just wondering why
ifw with the default/standard compression setting is 50% slower in imaging
than the competitors.

As you suggested I found that 'Enhanced Speed - A' is realy a runner which
can compete with the market in speed whether the compression ratio can't.

Images are 8-10% larger as with standard compression and compared to
competitors.

Looking at the available compression settings in ifw shows, that the most
are focused on compression ratio but only two on speed.

So a realy fast algorithm with good compression that can compete with the
market is currently missing in ifw (?).

G.

jelson
Posts: 5
Joined: Mon Nov 04, 2013 6:45 am

Re: IFW Imaging Speed

Post by jelson »

TeraByte Support(PP) wrote:
> Standard compression is not the best to test with when trying to check
> drive throughput. A better test would be using no compression (None) or
> Enhanced Speed - A (usually results in a backup image only slightly larger
> than Standard and is much faster).

Thank you. It makes an amazing difference simply switching from default (Std) to Enhanced Speed A (ESA).

IFW (Std) 117,100,640 sectors backed up in 18:07 -- Image size = 38.8 GB -- BU speed ==> 52.7 MB/s -- (Validate: 5:58)
IFW (ESA) 117,578,504 sectors backed up in 8:36 -- Image size = 36.6 GB -- BU speed ==> 111.3 MB/s -- (Validate: 4:26)

The latter result is definitely competitive... fast enough for me.

You mentioned some other settings in a later post:

When using PHYLock, both Disable Throttling on Reads / Writes should be ticked?

And you mentioned a "/po" option ... that enables caching for fast drives...? SSDs?
That and the other options mentioned /iocf & /iobs all have to be accessed via cmd line, correct?

Finally, backup speed is 1 concern, but my top priority is reliability. Any suggestions/considerations about that?
TeraByte Support
Posts: 3629
Joined: Thu May 05, 2011 10:37 pm

Re: IFW Imaging Speed

Post by TeraByte Support »

See below.

"jelson" wrote in message news:9069@public.image...

TeraByte Support(PP) wrote:
> Standard compression is not the best to test with when trying to check
> drive throughput. A better test would be using no compression (None) or
> Enhanced Speed - A (usually results in a backup image only slightly larger
> than Standard and is much faster).

Thank you. It makes an amazing difference simply switching from default
(Std) to Enhanced Speed A (ESA).

IFW (Std) 117,100,640 sectors backed up in 18:07 -- Image size = 38.8
GB -- BU speed ==> 52.7 MB/s -- (Validate: 5:58)
IFW (ESA) 117,578,504 sectors backed up in 8:36 -- Image size = 36.6
GB -- BU speed ==> 111.3 MB/s -- (Validate: 4:26)

The latter result is definitely competitive... fast enough for me.

You mentioned some other settings in a later post:

When using PHYLock, both Disable Throttling on Reads / Writes should be
ticked?

++++
"Should" isn't the correct word. If throttling is enabled, it keeps
applications more responsive at the cost of a slower backup. But you can
use VSS and all those options to get the fastest backup (less data to backup
and doesn't need throttling).
++++

And you mentioned a "/po" option ... that enables caching for fast
drives...? SSDs?

++++
Manual has details. It's actually /po:0 that would help on some "removable"
drives. If it's not removable, the default is already cached.
++++

That and the other options mentioned /iocf & /iobs all have to be accessed
via cmd line, correct?

++++
Yes or via the ifw.ini file (manual has details).
++++

Finally, backup speed is 1 concern, but my top priority is reliability. Any
suggestions/considerations about that?

++++
That's always enabled and in there - byte-for-byte does the most checking
++++


Post Reply