[DRBD-user] Fw: DRBD SSD's and TRIM

Håkan Engblom zyber_cynic at hotmail.com
Tue Nov 15 13:36:54 CET 2011

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


Hi,
 
> I believe that this is no problem at all, as it writes only zeros (if the 
> source disc is "really" empty).
Unfortunately, in our product, I cannot assume that the source disk is "filled with zeroes" in those parts of the disk that is "free".
 
> Why I come to this conclusion? On SSD discs without trim support and no trim 
> utility from the vendor, it's a well known technique to use SecureErase and 
> write zeros all over the disc, to restore factory default 
> settings/performance for such discs. (and this works!)
Thanks for this info, I was not 100 % sure this worked fine. However setting a disk to factory default by writing zeroes to the entire disk requires taking that disk out of service ofcourse. 
This is also a problem in the product I work with. So the options are limited.
There are specific situations in our products life-cycle where this is possible to do, and there we already do it. 
So maybe I can use this a "fallback remedy" to recover SSD perfromance.
 
br Håkan
 

> From: christoph at iway.ch
> To: drbd-user at lists.linbit.com
> Date: Tue, 15 Nov 2011 12:21:29 +0100
> Subject: [DRBD-user] Fw: DRBD SSD's and TRIM
> 
> Hello Håkan
> 
> >> secondary host/SSD. This is a bad thing to do with an SSD's since it
> >> will conclude every block on the disk is occupied by "useful" data, and
> >> it will have a large negative impact on the SSD's wear leveling
> >> algorithm. Can somebody confim this conclusion ?
> > Well, you can easily avoid the initial sync; as "unused" blocks will be
> > returned as zeroes it won't cause "different" data on both sides.
> 
> I believe that this is no problem at all, as it writes only zeros (if the 
> source disc is "really" empty). And any SSD should not taint the flash cells 
> if it sees that a complete erase block (I believe this is between 2MB and 
> 8MB on todays SSDs?) only consists of zeros.
> 
> Why I come to this conclusion? On SSD discs without trim support and no trim 
> utility from the vendor, it's a well known technique to use SecureErase and 
> write zeros all over the disc, to restore factory default 
> settings/performance for such discs. (and this works!)
> 
> regards
> Christoph
> 
> We're using SSDs in production systems for ~3 years now and had only 
> positive experiences so far. (we have way over hundred SSDs already deployed 
> in servers) 
> 
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20111115/8674076c/attachment.htm>


More information about the drbd-user mailing list