From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 67EEA5625EB for ; Wed, 9 Sep 2026 13:13:14 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788959596; cv=none; b=bosBna/yl+9z177OA243W1JU23lMZU5phnNo+03TmKmhax5IX+vT4dlJ7fJyvmn79qkqxytUbFqa0A1CDBdYBOJ0mJB3L0Re43aTB3gnFFZrIqtG2/Q58LIwtPJFQkwuR/fArij9xACkDH/eB8L/gcl9VWtB3O/Cx6dbHsnrBII= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788959596; c=relaxed/simple; bh=N34rlaLgq4ZPYDXg75ZzZJ7KLjgtuCGG0u2AYm57yk8=; h=Message-ID:Subject:From:To:Cc:Date:In-Reply-To:References: Content-Type:MIME-Version; b=vFJegO9F1nZ5X3raggIVPgm13MQCvb/JmgtOt3Yy676+37WyNtuHwmfPfe7lZ6FFTYQezv8tkpEDw3e2P/1lEtah9Amx+jOUMGVtsPensDf9hi/Z/3kdAKfRuUUCiXlZrdb+g3v27iiYSr87/UTsJYArptd2xLkZSmgJDERzp/o= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=ikOnEgXX; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="ikOnEgXX" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 865E21F00A3A; Wed, 9 Sep 2026 13:13:13 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1788959594; bh=oTcd4Mjly4VIFu5KN2Wh0gE7C53D9T1+WRZXaSEjBlM=; h=Subject:From:To:Cc:Date:In-Reply-To:References; b=ikOnEgXXxZeznbV9hpDHoxmGU4oVTPgp2GQvrnSM5+q2pvpAPWfm0VY68fC2Q0tvw kGLbCnufAMyqQnOf+SuAC2Oyv/Q91WgWdj7xo68CtCuMv9RgYTjvWCEXcv0piT6Xec VQ/uZqTi+826pxXiAJyRBon+JM6HcHHWwcIriG0ieytXd0eG1sjXy9WtUcj4xiGcfS WeFp1cemOadTvmqvXLakVtAPhe0WZEWeW2t8wtFf20AlE51pxvziahHO/JCgDu27Fj ntgD3Ses6EKDJFE54HDBQlvJ1jpYyAqPkn4udskjSvN7yNGCE4/Io+N0h00YFfX5ay mwq2BWOYd+LHw== Message-ID: <6b4764058582acb5a71d83fa9639df10b4c2f717.camel@kernel.org> Subject: Re: NFS delegations behavior analysis From: Jeff Layton To: Piyush Sachdeva , Anna Schumaker , Tigran Mkrtchyan Cc: linux-nfs , Chuck Lever , Trond Myklebust , sfrench@samba.org, sprasad@microsoft.com, vaibsharma@microsoft.com Date: Wed, 09 Sep 2026 09:13:11 -0400 In-Reply-To: References: <0b39c1e01a92f99fe456c76523ec7f3aa5dc1a81.camel@kernel.org> <455619640.1622514.1782212671358.JavaMail.zimbra@desy.de> Autocrypt: addr=jlayton@kernel.org; prefer-encrypt=mutual; keydata=mQINBE6V0TwBEADXhJg7s8wFDwBMEvn0qyhAnzFLTOCHooMZyx7XO7dAiIhDSi7G1NPxw n8jdFUQMCR/GlpozMFlSFiZXiObE7sef9rTtM68ukUyZM4pJ9l0KjQNgDJ6Fr342Htkjxu/kFV1Wv egyjnSsFt7EGoDjdKqr1TS9syJYFjagYtvWk/UfHlW09X+jOh4vYtfX7iYSx/NfqV3W1D7EDi0PqV T2h6v8i8YqsATFPwO4nuiTmL6I40ZofxVd+9wdRI4Db8yUNA4ZSP2nqLcLtFjClYRBoJvRWvsv4lm 0OX6MYPtv76hka8lW4mnRmZqqx3UtfHX/hF/zH24Gj7A6sYKYLCU3YrI2Ogiu7/ksKcl7goQjpvtV YrOOI5VGLHge0awt7bhMCTM9KAfPc+xL/ZxAMVWd3NCk5SamL2cE99UWgtvNOIYU8m6EjTLhsj8sn VluJH0/RcxEeFbnSaswVChNSGa7mXJrTR22lRL6ZPjdMgS2Km90haWPRc8Wolcz07Y2se0xpGVLEQ cDEsvv5IMmeMe1/qLZ6NaVkNuL3WOXvxaVT9USW1+/SGipO2IpKJjeDZfehlB/kpfF24+RrK+seQf CBYyUE8QJpvTZyfUHNYldXlrjO6n5MdOempLqWpfOmcGkwnyNRBR46g/jf8KnPRwXs509yAqDB6sE LZH+yWr9LQZEwARAQABtCVKZWZmIExheXRvbiA8amxheXRvbkBwb29jaGllcmVkcy5uZXQ+iQI7BB MBAgAlAhsDBgsJCAcDAgYVCAIJCgsEFgIDAQIeAQIXgAUCTpXWPAIZAQAKCRAADmhBGVaCFc65D/4 gBLNMHopQYgG/9RIM3kgFCCQV0pLv0hcg1cjr+bPI5f1PzJoOVi9s0wBDHwp8+vtHgYhM54yt43uI 7Htij0RHFL5eFqoVT4TSfAg2qlvNemJEOY0e4daljjmZM7UtmpGs9NN0r9r50W82eb5Kw5bc/r0km R/arUS2st+ecRsCnwAOj6HiURwIgfDMHGPtSkoPpu3DDp/cjcYUg3HaOJuTjtGHFH963B+f+hyQ2B rQZBBE76ErgTDJ2Db9Ey0kw7VEZ4I2nnVUY9B5dE2pJFVO5HJBMp30fUGKvwaKqYCU2iAKxdmJXRI ONb7dSde8LqZahuunPDMZyMA5+mkQl7kpIpR6kVDIiqmxzRuPeiMP7O2FCUlS2DnJnRVrHmCljLkZ Wf7ZUA22wJpepBligemtSRSbqCyZ3B48zJ8g5B8xLEntPo/NknSJaYRvfEQqGxgk5kkNWMIMDkfQO lDSXZvoxqU9wFH/9jTv1/6p8dHeGM0BsbBLMqQaqnWiVt5mG92E1zkOW69LnoozE6Le+12DsNW7Rj iR5K+27MObjXEYIW7FIvNN/TQ6U1EOsdxwB8o//Yfc3p2QqPr5uS93SDDan5ehH59BnHpguTc27Xi QQZ9EGiieCUx6Zh2ze3X2UW9YNzE15uKwkkuEIj60NvQRmEDfweYfOfPVOueC+iFifbQgSmVmZiBM YXl0b24gPGpsYXl0b25AcmVkaGF0LmNvbT6JAjgEEwECACIFAk6V0q0CGwMGCwkIBwMCBhUIAgkKC wQWAgMBAh4BAheAAAoJEAAOaEEZVoIViKUQALpvsacTMWWOd7SlPFzIYy2/fjvKlfB/Xs4YdNcf9q LqF+lk2RBUHdR/dGwZpvw/OLmnZ8TryDo2zXVJNWEEUFNc7wQpl3i78r6UU/GUY/RQmOgPhs3epQC 3PMJj4xFx+VuVcf/MXgDDdBUHaCTT793hyBeDbQuciARDJAW24Q1RCmjcwWIV/pgrlFa4lAXsmhoa c8UPc82Ijrs6ivlTweFf16VBc4nSLX5FB3ls7S5noRhm5/Zsd4PGPgIHgCZcPgkAnU1S/A/rSqf3F LpU+CbVBDvlVAnOq9gfNF+QiTlOHdZVIe4gEYAU3CUjbleywQqV02BKxPVM0C5/oVjMVx3bri75n1 TkBYGmqAXy9usCkHIsG5CBHmphv9MHmqMZQVsxvCzfnI5IO1+7MoloeeW/lxuyd0pU88dZsV/riHw 87i2GJUJtVlMl5IGBNFpqoNUoqmvRfEMeXhy/kUX4Xc03I1coZIgmwLmCSXwx9MaCPFzV/dOOrju2 xjO+2sYyB5BNtxRqUEyXglpujFZqJxxau7E0eXoYgoY9gtFGsspzFkVNntamVXEWVVgzJJr/EWW0y +jNd54MfPRqH+eCGuqlnNLktSAVz1MvVRY1dxUltSlDZT7P2bUoMorIPu8p7ZCg9dyX1+9T6Muc5d Hxf/BBP/ir+3e8JTFQBFOiLNdFtB9KZWZmIExheXRvbiA8amxheXRvbkBzYW1iYS5vcmc+iQI4BBM BAgAiBQJOldK9AhsDBgsJCAcDAgYVCAIJCgsEFgIDAQIeAQIXgAAKCRAADmhBGVaCFWgWD/0ZRi4h N9FK2BdQs9RwNnFZUr7JidAWfCrs37XrA/56olQl3ojn0fQtrP4DbTmCuh0SfMijB24psy1GnkPep naQ6VRf7Dxg/Y8muZELSOtsv2CKt3/02J1BBitrkkqmHyni5fLLYYg6fub0T/8Kwo1qGPdu1hx2BQ RERYtQ/S5d/T0cACdlzi6w8rs5f09hU9Tu4qV1JLKmBTgUWKN969HPRkxiojLQziHVyM/weR5Reu6 FZVNuVBGqBD+sfk/c98VJHjsQhYJijcsmgMb1NohAzwrBKcSGKOWJToGEO/1RkIN8tqGnYNp2G+aR 685D0chgTl1WzPRM6mFG1+n2b2RR95DxumKVpwBwdLPoCkI24JkeDJ7lXSe3uFWISstFGt0HL8Eew P8RuGC8s5h7Ct91HMNQTbjgA+Vi1foWUVXpEintAKgoywaIDlJfTZIl6Ew8ETN/7DLy8bXYgq0Xzh aKg3CnOUuGQV5/nl4OAX/3jocT5Cz/OtAiNYj5mLPeL5z2ZszjoCAH6caqsF2oLyAnLqRgDgR+wTQ T6gMhr2IRsl+cp8gPHBwQ4uZMb+X00c/Amm9VfviT+BI7B66cnC7Zv6Gvmtu2rEjWDGWPqUgccB7h dMKnKDthkA227/82tYoFiFMb/NwtgGrn5n2vwJyKN6SEoygGrNt0SI84y6hEVbQlSmVmZiBMYXl0b 24gPGpsYXl0b25AcHJpbWFyeWRhdGEuY29tPokCOQQTAQIAIwUCU4xmKQIbAwcLCQgHAwIBBhUIAg kKCwQWAgMBAh4BAheAAAoJEAAOaEEZVoIV1H0P/j4OUTwFd7BBbpoSp695qb6HqCzWMuExsp8nZjr uymMaeZbGr3OWMNEXRI1FWNHMtcMHWLP/RaDqCJil28proO+PQ/yPhsr2QqJcW4nr91tBrv/MqItu AXLYlsgXqp4BxLP67bzRJ1Bd2x0bWXurpEXY//VBOLnODqThGEcL7jouwjmnRh9FTKZfBDpFRaEfD FOXIfAkMKBa/c9TQwRpx2DPsl3eFWVCNuNGKeGsirLqCxUg5kWTxEorROppz9oU4HPicL6rRH22Ce 6nOAON2vHvhkUuO3GbffhrcsPD4DaYup4ic+DxWm+DaSSRJ+e1yJvwi6NmQ9P9UAuLG93S2MdNNbo sZ9P8k2mTOVKMc+GooI9Ve/vH8unwitwo7ORMVXhJeU6Q0X7zf3SjwDq2lBhn1DSuTsn2DbsNTiDv qrAaCvbsTsw+SZRwF85eG67eAwouYk+dnKmp1q57LDKMyzysij2oDKbcBlwB/TeX16p8+LxECv51a sjS9TInnipssssUDrHIvoTTXWcz7Y5wIngxDFwT8rPY3EggzLGfK5Zx2Q5S/N0FfmADmKknG/D8qG IcJE574D956tiUDKN4I+/g125ORR1v7bP+OIaayAvq17RP+qcAqkxc0x8iCYVCYDouDyNvWPGRhbL UO7mlBpjW9jK9e2fvZY9iw3QzIPGKtClKZWZmIExheXRvbiA8amVmZi5sYXl0b25AcHJpbWFyeWRh dGEuY29tPokCOQQTAQIAIwUCU4xmUAIbAwcLCQgHAwIBBhUIAgkKCwQWAgMBAh4BAheAAAoJEAAOa EEZVoIVzJoQALFCS6n/FHQS+hIzHIb56JbokhK0AFqoLVzLKzrnaeXhE5isWcVg0eoV2oTScIwUSU apy94if69tnUo4Q7YNt8/6yFM6hwZAxFjOXR0ciGE3Q+Z1zi49Ox51yjGMQGxlakV9ep4sV/d5a50 M+LFTmYSAFp6HY23JN9PkjVJC4PUv5DYRbOZ6Y1+TfXKBAewMVqtwT1Y+LPlfmI8dbbbuUX/kKZ5d dhV2736fgyfpslvJKYl0YifUOVy4D1G/oSycyHkJG78OvX4JKcf2kKzVvg7/Rnv+AueCfFQ6nGwPn 0P91I7TEOC4XfZ6a1K3uTp4fPPs1Wn75X7K8lzJP/p8lme40uqwAyBjk+IA5VGd+CVRiyJTpGZwA0 jwSYLyXboX+Dqm9pSYzmC9+/AE7lIgpWj+3iNisp1SWtHc4pdtQ5EU2SEz8yKvDbD0lNDbv4ljI7e flPsvN6vOrxz24mCliEco5DwhpaaSnzWnbAPXhQDWb/lUgs/JNk8dtwmvWnqCwRqElMLVisAbJmC0 BhZ/Ab4sph3EaiZfdXKhiQqSGdK4La3OTJOJYZphPdGgnkvDV9Pl1QZ0ijXQrVIy3zd6VCNaKYq7B AKidn5g/2Q8oio9Tf4XfdZ9dtwcB+bwDJFgvvDYaZ5bI3ln4V3EyW5i2NfXazz/GA/I/ZtbsigCFc 8ftCBKZWZmIExheXRvbiA8amxheXRvbkBrZXJuZWwub3JnPokCOAQTAQIAIgUCWe8u6AIbAwYLCQg HAwIGFQgCCQoLBBYCAwECHgECF4AACgkQAA5oQRlWghUuCg/+Lb/xGxZD2Q1oJVAE37uW308UpVSD 2tAMJUvFTdDbfe3zKlPDTuVsyNsALBGclPLagJ5ZTP+Vp2irAN9uwBuacBOTtmOdz4ZN2tdvNgozz uxp4CHBDVzAslUi2idy+xpsp47DWPxYFIRP3M8QG/aNW052LaPc0cedYxp8+9eiVUNpxF4SiU4i9J DfX/sn9XcfoVZIxMpCRE750zvJvcCUz9HojsrMQ1NFc7MFT1z3MOW2/RlzPcog7xvR5ENPH19ojRD CHqumUHRry+RF0lH00clzX/W8OrQJZtoBPXv9ahka/Vp7kEulcBJr1cH5Wz/WprhsIM7U9pse1f1g Yy9YbXtWctUz8uvDR7shsQxAhX3qO7DilMtuGo1v97I/Kx4gXQ52syh/w6EBny71CZrOgD6kJwPVV AaM1LRC28muq91WCFhs/nzHozpbzcheyGtMUI2Ao4K6mnY+3zIuXPygZMFr9KXE6fF7HzKxKuZMJO aEZCiDOq0anx6FmOzs5E6Jqdpo/mtI8beK+BE7Va6ni7YrQlnT0i3vaTVMTiCThbqsB20VrbMjlhp f8lfK1XVNbRq/R7GZ9zHESlsa35ha60yd/j3pu5hT2xyy8krV8vGhHvnJ1XRMJBAB/UYb6FyC7S+m QZIQXVeAA+smfTT0tDrisj1U5x6ZB9b3nBg65kc= Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable User-Agent: Evolution 3.60.2 (3.60.2-1.fc44) Precedence: bulk X-Mailing-List: linux-nfs@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 On Wed, 2026-09-09 at 15:40 +0530, Piyush Sachdeva wrote: > "Anna Schumaker" writes: >=20 > > On Tue, Jun 23, 2026, at 7:04 AM, Mkrtchyan, Tigran wrote: > > > ----- Original Message ----- > > > > From: "Jeff Layton" > > > > To: "Piyush Sachdeva" , "linux-nfs" , "Chuck Lever" , > > > > "trondmy" , sfrench@samba.org, sprasad@microsof= t.com > > > > Cc: vaibsharma@microsoft.com > > > > Sent: Tuesday, 23 June, 2026 12:50:16 > > > > Subject: Re: NFS delegations behavior analysis > > >=20 > > > > On Tue, 2026-06-23 at 15:31 +0530, Piyush Sachdeva wrote: > > > > > Hi, > > > > > Lately I have been running micro benchmarks around the `ls` comma= nd and > > > > > reading through the code documentation of the NFS client to bette= r > > > > > understand the client side caching behavior with and without > > > > > delegations. > > > > >=20 > > > > > Understanding so far: > > > > > Delegations (both file and directory) are granted by the server t= o the > > > > > client, indefinitely (until revoked or under the watermark) to ca= che > > > > > attributes. The caching of data is a result of the attribute > > > > > cache. Hence forth, a directory delegation will cache the directo= ry > > > > > attributes and the names of the files in the directory, and a fil= e > > > > > delegation will cache the attributes of the file and the file dat= a. > > > > >=20 > > > > > Workload run: > > > > > I focused on the 2 workloads below, doing 2 passes of a large fla= t > > > > > directory (with close to 100K files) - > > > > > a cold pass, and warm pass using the cache from the cold pass: > > > > > - lslr - ls -lR on both runs > > > > > - lsmix - ls -R (cold) and then ls -lR (warm) > > > > >=20 > > > > > I also played with the rdirplus behavior using both the default > > > > > heuristic behavior and the `rdirplus=3Dforce` set at mount time. > > > > >=20 > > > > > Numbers: > > > > > actimeo=3D5s, rdirplus=3Dforce, ACLs off, flat_dir > > > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > > > > >=20 > > > > > | LSLR | LSMIX > > > > > | (ls -lR cold / warm) | (p1 ls -R / p2 ls -lR= ) > > > > > Operation | flat cold | flat warm | flat p1 | flat p= 2 > > > > > -----------------+-------------+-----------+-------------+-------= -- > > > > > READDIR calls | 27 | 0 | 27 | 0 > > > > > READDIR recv B | 23,603,024 | 0 | 23,603,024 | 0 > > > > > call type | readdirplus | -- | readdirplus | -- > > > > > LOOKUP | 1 | 0 | 1 | 0 > > > > > GETATTR | 3 | 100,000 | 2 | 100,00= 1 > > > > > ACCESS | 2 | 0 | 2 | 0 > > > > > -----------------+-------------+-----------+-------------+-------= -- > > > > > Elapsed (age) | ~14 s | ~62 s | ~16 s | ~63 s > > > > >=20 > > > > >=20 > > > > > Observations: > > > > > When doing `ls` or `ls -l` on a directory, due to the open(2) on = the > > > > > directory, the client gets a directory delegation - caching the > > > > > directory attributes and file names. However, as we don't have fi= le > > > > > delegations due to no open(2) calls to any of the files. Hencefor= th, > > > > > the cache of file attributes is governed by `actimeo`. > > > > > Now here is the interesting bit, if the next `ls -l` is issued af= ter > > > > > the `actimeo`, a massive GETATTR storm hits the server, doing sta= t() > > > > > calls for every file in the directory. As a result, the performan= ce of > > > > > this warm `ls -l` run ends up being worse than the cold pass. I a= m > > > > > guessing this is most likely due to the compounded "rdirplus" bei= ng more > > > > > efficient than stat() calls. > > > > >=20 > > > > >=20 > > > > > Proposal: > > > > > For large directories, this ends up being a massive problem, taki= ng 1-2 > > > > > minutes when enumerating a directory on the warm passes. > > > > > - An easier way to tackle this could be to do a rdirplus=3D[auto = | forced] > > > > > instead of issuing the stat(2) storm to the server: When the cl= ient > > > > > notices that there are cache misses, which would be the case of= file > > > > > attributes, instead of fetching file names from the directory-d= elegation > > > > > cache and attributes from GETATTR, the client does a READDIRPLU= S to > > > > > the server, nonetheless. > > > > > - A more tedious would be the to cache file attributes as well, a= s a part > > > > > of the directory delegation. This would end up requiring a chan= ge in the > > > > > NFS protocol spec though. > > > > > - Bulk GETATTR calls: I am uncertain of the feasibility of this, = but > > > > > what if, the client could do 1 GETATTR call for getting attribu= tes > > > > > for multiple files. > > > >=20 > > > >=20 > > > > ls is such a hard workload to get right, because we don't really ge= t an > > >=20 > > > 100% agree. And there were a couple of attempts to address this issue > > > (second ls that is slow). > > >=20 > > > > indication in the kernel of what userland's intentions are. It's > > > > basically a readdir() call followed by a bunch of stat()'s, but at = the > > > > point where we're getting the readdir() call, we don't know if user= land > > > > intends to stat() those files or not. We have to make a guess about > > > > that intention. > > > >=20 > > > > In this case, it sounds like the directory cache was valid, so the > > > > client decided it didn't need to do a READDIR at all, but the > > > > individual files had caches that timed out. > > > >=20 > > > > So imagine you're the kernel client and have been given that second > > > > readdir() call: Why should you decide to do a READDIRPLUS at that p= oint > > > > instead of a regular READDIR? > > >=20 > > > May we need some kind of client-side heuristics, like on the server s= ide > > > for open-delegations, where after seeing some `stats` for files in th= e > > > In the same directory, the client will decide to switch to READDIR (v= 4) > > > to get all attributes in one go. > >=20 > > We do something like that already. I don't think we'll ever have a read= dir > > plus heuristic that makes everybody happy without userspace somehow tel= ling > > us their intentions. I know a statx-based readdir plus system call prob= ably > > sounds crazy, but something like that would go a long way to take all t= he > > guesswork out of things on our end. > >=20 > > Anna > >=20 > > >=20 > > > Best regards, > > > Tigran. > > >=20 > > > > -- > > > > Jeff Layton > > >=20 > > > Attachments: > > > * smime.p7s >=20 > Hi, > I believe the problem from this discussion has been the cost associated > with GETATTRs during directory enumerations, compounded by the fact that > the kernel can't know whether more GETATTRs on the same directory are > coming. >=20 > Since the readdirplus heuristic rework in v5.18, plaing READDIRs > fetching the file type in v6.8, and introduction of rdirplus=3Dforce in > v6.14, the client has become quite performant especially for the > ls-style workloads. find and du, still suffer from GETATTR storms for > large-enough directories, as they enumerate the directory once then > stat() every child. Note that nfs_getattr() already feeds the > cache_hit/cache_miss counters (since v4.10), but that signal lives on > the open-directory context and is consumed only inside nfs_readdir(). > This is where the performance hurts the most. >=20 > As a benchmark, I prototyped a solution for older kernels (v4.18, v5.14, > v6.4-based distros). An eBPF-based detector recognizes GETATTR/LOOKUP > storms on large flat directories and triggers a userspace helper to > enumerate the directory (READDIR(PLUS), start -> EOF). The returned > entries refresh the attribute/dcache so subsequent GETATTR/LOOKUP calls > are served from cache. We measured directory-enumeration time and > GETATTR/LOOKUP counts drop by ~50=E2=80=9380x on large directories, for l= s, > find, and du. >=20 > With that as evidence, I'd like to propose exploring two in-kernel > directions and get your opinion on feasibility and desirability: > 1. Extend the existing getattr miss-tracking: when cache_misses on a > directory cross a threshold, trigger a deferred, background > READDIRPLUS to refresh that directory's child attributes. This gives > the find / du case a consumer for the miss signal it already > generates. Tough call. This adds load to the server for attributes that will never be used. My gut feeling is that that might help some workloads, but hurt others. > 2. A batched-GETATTR path so tools like du /find can request many > children's attributes in fewer round-trips. I recognize there's no > batched-stat() interface today and that this spans userspace and the > kernel, so I'm raising it mainly to ask whether a metadata-batching > interface (or wiring stat readahead to an NFSv4 multi-GETATTR > compound) is something the community would consider. >=20 > I'd appreciate your thoughts on both, particularly whether (1) is best > framed as an extension of the v5.18 heuristic. #2 sounds like a mess to get right.... What might be more reasonable is adding a real readdirplus() style- system call, like Anna mentioned. Something where userland can do a readdir() and pass down a statx() mask to tell the kernel what attributes it's interested in, and get back a list of filenames and their inodes' attributes in one shot. That too might not be accepted upstream (it has been attempted before), but that seems like it would be a fairly straightforward thing for userland, and would work well for "ls -l" and the like. --=20 Jeff Layton