Soren Harward on 16 Apr 2015 11:44:52 -0700

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] swappiness and ssd

On Thu, Apr 16, 2015 at 2:09 PM, Rich Freeman <> wrote:
> Those kinds of stats assume certain kinds of use patterns.  Lots of
> internal writes to a swap file may or may not wear evenly across the
> entire SSD memory.
> I'm not saying it is a recipe for disaster, but I'd be careful in
> extrapolating benchmarks without understanding how they were
> collected.

True, they pounded the whole drive instead of a single partition.  So
having heavy swap activity on a single partition may wear out those
blocks more quickly if the drive doesn't distribute the data cleverly.
On the other hand, SSDs typically have a couple GB of spare blocks to
replace the broken ones.  So if you had 16GB of spare blocks (typical
for a 240GB drive), and a 8GB swap partition, you could destroy all
blocks in the partition twice over before running into problems.  And
with 8GB of swap, you're not going to be filling up that partition too

I acknowledge that it's possible to wear out a consumer-grade SSD.
I'm just pointing out you're highly unlikely to do so without
intentionally abusing the drive.

Soren Harward
Philadelphia Linux Users Group         --
Announcements -
General Discussion  --