Page 1 of 1

Two questions about L2 Gather Algorithm

Posted: Sun Sep 05, 2021 10:34 am
by RobF99
1. Does L2 Gather gather the data in an optimized way e.g. gathers data in sequential order according to cluster number?

2. Does Instant gather setting in any way interfere with file access if the drive is accessing data from the hard drive or is the data just written to L2 immediately as it is accessed from the hard drive? I tried testing this but could not come to any meaningful conclusion.

Thanks in advance.

Re: Two questions about L2 Gather Algorithm

Posted: Mon Sep 06, 2021 12:10 pm
by Support
RobF99 wrote: Sun Sep 05, 2021 10:34 am 1. Does L2 Gather gather the data in an optimized way e.g. gathers data in sequential order according to cluster number?
Yes.
RobF99 wrote: Sun Sep 05, 2021 10:34 am 2. Does Instant gather setting in any way interfere with file access if the drive is accessing data from the hard drive or is the data just written to L2 immediately as it is accessed from the hard drive? I tried testing this but could not come to any meaningful conclusion.
Cache integrity/synchronization has already been considered to ensure cached data is correct and updated.

Re: Two questions about L2 Gather Algorithm

Posted: Mon Sep 06, 2021 7:10 pm
by RobF99
Thanks. I should have been more clear on question 2. I meant file read performance. Does Instant Gather slow down read performance at all from the hard drive if data is not in L2 since it will instantaneously read HDD and write to L2.

Re: Two questions about L2 Gather Algorithm

Posted: Tue Sep 07, 2021 1:57 am
by Support
Processing read requests and populating L2 data are still separated. The "Instant" option ignores the device busy status and writing to L2 will almost instantly triggered after a fresh read request. So for a single read request, the performance will not be affected. However, if reading requests are quite heavy, the whole performance might be affected.