1 |
On Sunday, 5 April 2020 18:54:25 BST tuxic@××××××.de wrote: |
2 |
> On 04/05 10:33, Mark Knecht wrote: |
3 |
> > On Sun, Apr 5, 2020 at 10:13 AM <tuxic@××××××.de> wrote: |
4 |
> > > Hi, |
5 |
> > > |
6 |
> > > currentlu I am preparing a new Gentoo Linux by compiling all |
7 |
> > > the application I had on my old system. |
8 |
> > > |
9 |
> > > Due to delivery problems (corona) my SSD was delivered today |
10 |
> > > (or yesterday...it depends...;) . |
11 |
> > > |
12 |
> > > When the whole compilation has finished and the system boots it |
13 |
> > > needs to be transfered to the SSD. |
14 |
> > > |
15 |
> > > The SSD has a heat spreader...so it gets hot, when used. |
16 |
> > > |
17 |
> > > Is it wise to copy the whole root system to the SSD in one go |
18 |
> > > in respect to a not so healthy heat increase? |
19 |
> > > |
20 |
> > > And if not...how can I copy the root system in portions |
21 |
> > > to the SSD and do not miss anything? |
22 |
> > > |
23 |
> > > Are there SDD-friendly and SSD-unfriendlu methods of copying |
24 |
> > > greater chunks of data to a SSD (rsync, tar-pipe, cp....)? |
25 |
> > > What is recommended here? |
26 |
> > > |
27 |
> > > Thanks a lot for any help for a SSD newbie in advance! |
28 |
> > > |
29 |
> > > Cheers! And stay heathy! |
30 |
> > > Meino |
31 |
> > |
32 |
> > Just my 2 cents... |
33 |
> > |
34 |
> > If the SSD cannot survive having data copied to it there's something |
35 |
> > seriously wrong with the drive. I don't think you should be overly worried |
36 |
> > about this but I do understand it's new technology so you want to be |
37 |
> > careful. Bravo for that. |
38 |
> > |
39 |
> > Possibly to ease your concerns a little bit use smartctl -a /dev/SSD and |
40 |
> > get to know your drive that way. You can most likely watch the drive temp |
41 |
> > as recorded by the drive. |
42 |
> > |
43 |
> > Best wishes, |
44 |
> > Mark |
45 |
> |
46 |
> Hi Mark, |
47 |
> |
48 |
> Yes, if a SSD could not survive writes, something is wrong with the |
49 |
> SSD. But that was not my point. |
50 |
> Copying about 100GB (roughly guessed) data in one go to the SSD is a |
51 |
> use case, which is not common. And therefore possibly not taken into |
52 |
> account by the company, which create that SSD. |
53 |
> SSDs can create noticeable heat (mine has a minimalistic heat |
54 |
> spreader therefore. Faster SSDs come with a substancial heatspreader). |
55 |
> |
56 |
> Smartctl will report problems when they are already there. |
57 |
> I want to prevent problems beforehand. |
58 |
> |
59 |
> So -- does copying about 100 GB creates so much heat in the sillicone |
60 |
> of the SSD, that it ages more than preferred? |
61 |
> |
62 |
> And if so, how can I prevent it by appluing other techniques to copy |
63 |
> the data? |
64 |
> See additional questions in my initial posting for that. |
65 |
|
66 |
With rsync you could add the option: |
67 |
|
68 |
--bwlimit=RATE limit socket I/O bandwidth |
69 |
|
70 |
and select a low enough bandwidth limit to keep the packets flowing at a low |
71 |
rate to remain cool enough for your liking. |
72 |
|
73 |
However, I'll echo what other have said about not worrying about these things. |
74 |
The OEMs must run some rudimentary tests to establish performance envelopes |
75 |
and keep everything tuned just so. |
76 |
|
77 |
Nevertheless, if you do not trust them to produce the best quality products, |
78 |
then we share something in common! LOL! |
79 |
|
80 |
In this case, you may want to open the PC case and use a desktop fan to keep |
81 |
the interior of the case even cooler than usual, during your copying process. |