From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matt Subject: Re: 2.6.37: Multi-second I/O latency while untarring Date: Fri, 11 Feb 2011 15:54:31 +0000 Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Cc: linux-btrfs@vger.kernel.org, linux-kernel@vger.kernel.org To: Andrew Lutomirski Return-path: In-Reply-To: List-ID: On Fri, Feb 11, 2011 at 3:08 PM, Andrew Lutomirski wrote= : > As I type this, I have an ssh process running that's dumping data int= o > a fifo at high speed (maybe 500Mbps) and a tar process that's > untarring from the same fifo onto btrfs. =A0The btrfs fs is mounted -= o > space_cache,compress. =A0This machine has 8GB ram, 8 logical cores, a= nd > a fast (i7-2600) CPU, so it's not an issue with the machine strugglin= g > under load. > > Every few tens of seconds, my system stalls for several seconds. > These stalls cause keyboard input to be lost, firefox to hang, etc. > > Setting tar's ionice priority to best effort / 7 or to idle makes no = difference. > > ionice idle and queue_depth =3D 1 on the disk (a slow 2TB WD) also ma= kes > no difference. > > max_sectors_kb =3D 64 in addition to the above doesn't help either. > > latencytop shows regular instances of 2-7 *second* latency, variously > in sync_page, start_transaction, btrfs_start_ordered_extent, and > do_get_write_access (from jbd2 on my ext4 root partition). > > echo 3 >drop_caches gave me 7 GB free RAM. =A0I still had stalls when > 4-5 GB were still free (so it shouldn't be a problem with important > pages being evicted). > > In case it matters, all of my partitions are on LVM on dm-crypt, but > this machine has AES-NI so the overhead from that should be minimal. > In fact, overall CPU usage is only about 10%. > > What gives? =A0I thought this stuff was supposed to be better on mode= rn kernels. > > --Andy > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs= " in > the body of a message to majordomo@vger.kernel.org > More majordomo info at =A0http://vger.kernel.org/majordomo-info.html > Hi Andrew, you could try the following patch to speed up dm-crypt: https://patchwork.kernel.org/patch/365542/ I'm using it on top of a highly-patched 2.6.37 kernel not sure if exactly that version was included in 2.6.38 there are some additional handles to speed up dm: e.g. PCRYCONFIG_CRYPTO_PCRYPT=3Dy Regards Matt