| To: | xfs@xxxxxxxxxxx |
|---|---|
| Subject: | Re: all your slabs are belong to ram ? |
| From: | "Johannes Truschnigg" <johannes.truschnigg@xxxxxxxxxxx> |
| Date: | Tue, 6 Oct 2015 18:16:46 +0200 |
| Cc: | "krautus@xxxxxxxxx" <krautus@xxxxxxxxx> |
| Delivered-to: | xfs@xxxxxxxxxxx |
| Dkim-signature: | v=1; a=rsa-sha256; c=simple/simple; d=geizhals.at; s=mail; t=1444148207; bh=FWWq+tmKKSA3f6Y1zYB2p5ff+KuBXpiXNa9TnJkVOQs=; h=Message-ID:In-Reply-To:References:Date:Subject:From:To:Cc: MIME-Version:Content-Type:Content-Transfer-Encoding; b=Lphc7Y+aKIOgFFMQsZXLs201TNlfvresH/JnKveT8vPpxBQLNNPj5AuhRU4ehkSBR 4Hyjcql8U1MxWl7nMtJm4OqnSEBUYDhahqk2VE0wthREdKAmv3RVl6UrnttG5T/orA FE85YegMwxMSjBJclX0Ae1TzH4m9Six4QUeOXRdU= |
| Dkim-signature: | v=1; a=rsa-sha256; c=simple/simple; d=geizhals.at; s=mail; t=1444148206; bh=FWWq+tmKKSA3f6Y1zYB2p5ff+KuBXpiXNa9TnJkVOQs=; h=Message-ID:In-Reply-To:References:Date:Subject:From:To:Cc: MIME-Version:Content-Type:Content-Transfer-Encoding; b=nzNibqu5Rs7GYyfZb50GMk21cEAwKpfsGyWbQLiC0inBCkUfyWjfiO7WPsj8wjC4/ iQcsOYdMzFJWqXy9NCH5VKTT3/XIqGww/tD7tfTqOjAYZ5JjjgXz+4ZNKggvEQmfjd 9n7me/nB4SEfu2nCQWWc5ZhhqSY82Ndozd9nh9QE= |
| Importance: | Normal |
| In-reply-to: | <20151006145038.GE10397@xxxxxxxxxxxxxxxx> |
| References: | <20151006162443.42f48ee7@linux> <20151006145038.GE10397@xxxxxxxxxxxxxxxx> |
| User-agent: | SquirrelMail/1.4.21 |
Am Di, 6.10.2015, 16:50 schrieb Darrick J. Wong: > On Tue, Oct 06, 2015 at 04:24:43PM +0200, krautus@xxxxxxxxx wrote: >> [...] >> So I'm asking you: >> 1. is there a way to force dentries and inodes to stay in ram ? >> 2. can I perhaps move dentries and inodes to a dedicated SSD ? >> >> I'm open to all possibilities, perhaps increase RAM ? >> Upgrade to Debian Jessie and 64 bit ? > > ISTR that kernel data such as slabs cannot live in highmem, which means > that > dentries and slab cannot live in highmem. A 32bit kernel sets up ~900M of > low > memory and ~15G of highmem, which is probably why the kernel has to evict > things and why you see such problems. > > A 64bit kernel sets up all the memory as lowmem, so the kernel can use all > the > memory for stuff like that. I'd give that a try first. A few years back, we solved pretty much that exact same problem by switching the kernel to amd64, with all of userspace remaining i386. You should definitely try this. -- Mit freundlichen Grüßen Johannes Truschnigg Senior System Administrator -- mailto:johannes.truschnigg@xxxxxxxxxxx (in dringenden Fällen bitte an info@xxxxxxxxxxx) Geizhals(R) - Preisvergleich Internet Services AG Obere Donaustrasse 63/2 A-1020 Wien Tel: +43 1 5811609/87 Fax: +43 1 5811609/55 http://geizhals.at => Preisvergleich für Österreich http://geizhals.de => Preisvergleich für Deutschland http://geizhals.eu => Preisvergleich EU-weit Handelsgericht Wien | FN 197241K | Firmensitz Wien |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: all your slabs are belong to ram ?, Darrick J. Wong |
|---|---|
| Next by Date: | Re: [PATCH 1/3] xfs_io: Add inode '-s' command to query physical size of largest inode, Brian Foster |
| Previous by Thread: | Re: all your slabs are belong to ram ?, Darrick J. Wong |
| Next by Thread: | Re: all your slabs are belong to ram ?, krautus@xxxxxxxxx |
| Indexes: | [Date] [Thread] [Top] [All Lists] |