From owner-freebsd-questions@freebsd.org Tue Jan 31 19:05:30 2017 Return-Path: Delivered-To: freebsd-questions@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 945EDCCA73A for ; Tue, 31 Jan 2017 19:05:30 +0000 (UTC) (envelope-from radek.krejca@starnet.cz) Received: from EXCHANGE.mail.starnet.cz (exchange.mail.starnet.cz [92.62.224.72]) (using TLSv1 with cipher AES128-SHA (128/128 bits)) (Client CN "Sites", Issuer "STARNET" (not verified)) by mx1.freebsd.org (Postfix) with ESMTPS id 279F4B62 for ; Tue, 31 Jan 2017 19:05:29 +0000 (UTC) (envelope-from radek.krejca@starnet.cz) Received: from EXCHANGE.mail.starnet.cz ([fe80::d017:9e72:12a5:7bb4]) by EXCHANGE.mail.starnet.cz ([fe80::d017:9e72:12a5:7bb4%14]) with mapi; Tue, 31 Jan 2017 20:05:20 +0100 From: =?iso-8859-2?Q?Radek_Krej=E8a?= To: "'freebsd-questions@freebsd.org'" Date: Tue, 31 Jan 2017 20:05:19 +0100 Subject: A lot of mbufs allocated Thread-Topic: A lot of mbufs allocated Thread-Index: AdJ79PhPDsdVbYeWR3anMGFAX+h9qw== Message-ID: Accept-Language: cs-CZ Content-Language: cs-CZ X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: cs-CZ Content-Type: text/plain; charset="iso-8859-2" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Jan 2017 19:05:30 -0000 Hello, I have problem with freezing FreeBSD boxes with PF on [zone: mbuf_cluster] = kern.ipc.nmbclusters limit reached last two months. But theres no evidence = (or I dont know, what I have to check yet) of huge traffic. If I can connec= t to server yet, I can see, that clusters are really exhausted: 206604/6681/213285 mbufs in use (current/cache/total) 206603/3697/210300/250866 mbuf clusters in use (current/cache/total/max) 206603/3640 mbuf+clusters out of packet secondary zone in use (current/cach= e) But why? How I can check what consume this memory? netstat -an, pfctlss -ss= shows nothing unusual. What I can check else? Another question - there are sins that freebsd 10.3 has different access to= manage memory clusters, as I can see on graphs over snmp. And last strange difference between 10.1 and 10.3 on graphs. First is the s= ame hardware, but freebsd 10.1, second 10.3.=20 https://postimg.org/image/5i7izm6rx/ https://postimg.org/image/j0efbwix9/ What else can I check? Thank you very much. Radek