Comments on: Back in the sandbox…ZFS flushing shenanigans revisted. http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/ thoughts & musings Sat, 24 Jul 2010 02:26:30 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Shenanigans with ZFS flushing and intelligent arrays… - Jason’s .plan http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/comment-page-1/#comment-712 Shenanigans with ZFS flushing and intelligent arrays… - Jason’s .plan Thu, 15 Jan 2009 19:29:13 +0000 #comment-712 [...] NOTE: ZFS has been enhanced to better address the situation described below by using ZFS configuration directives. This article is still accurate and provides decent background on the problem. However, an update has been posted with the newer, stronger, better way of resolving the problem: Back in the sandbox…ZFS flushing shenanigans revisited. [...] [...] NOTE: ZFS has been enhanced to better address the situation described below by using ZFS configuration directives. This article is still accurate and provides decent background on the problem. However, an update has been posted with the newer, stronger, better way of resolving the problem: Back in the sandbox…ZFS flushing shenanigans revisited. [...]

]]> By: Gerardo Diaz http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/comment-page-1/#comment-380 Gerardo Diaz Wed, 15 Oct 2008 08:14:07 +0000 #comment-380 If the option to not flush the cache is used, would that cause the filesystems memory (server memory not the arrays cache) buffers to not flush too? If the option to not flush the cache is used, would that cause the filesystems memory (server memory not the arrays cache) buffers to not flush too?

]]> By: Joe http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/comment-page-1/#comment-379 Joe Sun, 06 Jul 2008 03:56:57 +0000 #comment-379 I don't get it ...<br /> <br /> Your premise (in the prior article) is:<br /> <br /> * Tell your array to ignore ZFS' flush commands. This is pretty safe, and massively beneficial.<br /> <br /> The former option, is really a no go because it opens you up to losing data. The second option really works well and is darn safe. It ends up being safe because if ZFS is waiting for the write to complete, that means the write made it to the array, and if its in the array cache you're golden. Whether famine or flood or a loose power cable come, your array will get that write to the disk eventually. So its OK to have the array lie to ZFS and release ZFS almost immediately after the ZIL flush command executes. On our StorageTek FLX210 this took the idle latencies to 1ms and the heavy load latencies to 9ms. 9 bloody milliseconds! Our InnoDB problems disappeared like sand down a rat hole. <br /> <br /> How do you KNOW that the write will be made within 72 hours ?<br /> <br /> In the event of "flood or famine" it likely will not.<br /> <br /> How about I wire you a few million and you reply immediately that you received the cash, then prior to you actually writing the authentiction to your HD you get a tornado. I have my "proof" that I gave you the money but you don't have the money in your hand. I don't get it …

Your premise (in the prior article) is:

* Tell your array to ignore ZFS' flush commands. This is pretty safe, and massively beneficial.

The former option, is really a no go because it opens you up to losing data. The second option really works well and is darn safe. It ends up being safe because if ZFS is waiting for the write to complete, that means the write made it to the array, and if its in the array cache you're golden. Whether famine or flood or a loose power cable come, your array will get that write to the disk eventually. So its OK to have the array lie to ZFS and release ZFS almost immediately after the ZIL flush command executes. On our StorageTek FLX210 this took the idle latencies to 1ms and the heavy load latencies to 9ms. 9 bloody milliseconds! Our InnoDB problems disappeared like sand down a rat hole.

How do you KNOW that the write will be made within 72 hours ?

In the event of "flood or famine" it likely will not.

How about I wire you a few million and you reply immediately that you received the cash, then prior to you actually writing the authentiction to your HD you get a tornado. I have my "proof" that I gave you the money but you don't have the money in your hand.

]]> By: UB http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/comment-page-1/#comment-378 UB Thu, 22 May 2008 13:55:52 +0000 #comment-378 Works with a patched s10u3.<br /> Tip: Check Your kernel:<br /> echo "zfs_nocacheflush/D" | mdb -k Works with a patched s10u3.
Tip: Check Your kernel:
echo "zfs_nocacheflush/D" | mdb -k

]]> By: d.s.ivanov http://blogs.digitar.com/jjww/2007/10/back-in-the-sandbox-zfs-flushing-shenanigans-revisted/comment-page-1/#comment-377 d.s.ivanov Thu, 14 Feb 2008 18:46:09 +0000 #comment-377 It's work at Solaris 10u4. It's work at Solaris 10u4.

]]>