From owner-freebsd-hackers@freebsd.org Sat Feb 22 18:27:43 2020 Return-Path: Delivered-To: freebsd-hackers@mailman.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.nyi.freebsd.org (Postfix) with ESMTP id 31BBC248914 for ; Sat, 22 Feb 2020 18:27:43 +0000 (UTC) (envelope-from leres@freebsd.org) Received: from smtp.freebsd.org (smtp.freebsd.org [IPv6:2610:1c1:1:606c::24b:4]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) server-signature RSA-PSS (4096 bits) client-signature RSA-PSS (4096 bits) client-digest SHA256) (Client CN "smtp.freebsd.org", Issuer "Let's Encrypt Authority X3" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 48PxdV1Hnmz4HjV for ; Sat, 22 Feb 2020 18:27:42 +0000 (UTC) (envelope-from leres@freebsd.org) Received: from ice.alameda.xse.com (unknown [IPv6:2600:1700:a570:11f0:f2ad:4eff:fe0b:a065]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (Client did not present a certificate) (Authenticated sender: leres) by smtp.freebsd.org (Postfix) with ESMTPSA id C46E068AC for ; Sat, 22 Feb 2020 18:27:41 +0000 (UTC) (envelope-from leres@freebsd.org) To: FreeBSD Hackers From: Craig Leres Subject: Slow performance with one of two power supplies powered Message-ID: <8813b1e2-74af-5a13-adb9-36c4ed07cfcd@freebsd.org> Date: Sat, 22 Feb 2020 10:27:39 -0800 User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:68.0) Gecko/20100101 Thunderbird/68.5.0 MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 22 Feb 2020 18:27:43 -0000 I have two Supermicro X9SRi-F systems that are nearly identical. Both have dual power supplies fed from building and generator power. One is acting normally but the other does not have building power due to electrical work this weekend and has really poor performance. For example running wc on a 2.5M line file takes 2 seconds on the good system and 40 on the bad. Last time I ran into this I found the system was underclocking and "sysctl dev.cpu.0.freq" would report a lower than expected value. I solved this by running powerd with this config: powerd_enable="YES" powerd_flags="-a max -b max -n max" In this case dev.cpu.0.freq is reported as 2201 for both systems which seems correct for the 2.20GHz Intel E5-2660 processors in use. Running powerd does not change the reported frequency nor does it improve performance. I have another pair of X9SRi-F systems with the same power situation. They have 3.00GHz E5-2637 and the one running on one of two power supplies was reporting dev.cpu.0.freq as 1200 before running powerd and 3001 after. The single powered system also has lower performance, the wc test takes 14 seconds, the fully powered system takes < 0.5. How do I solve this? dmidecode reports the power supplies as PWS-406P-1R which is rated for 400W so I think they are beefy enough to run a single E5-2660, memory, and a couple of SSDs. Craig