::Gak Perlu Ragu Pake RC1 64bit untuk Mesin [Proxy] Anda::
-
kelebihannya …
bisa maksimalkan memory pisik sebanyak yg mainboard mampu
khan tahu sendiri squid/proxy itu cukup boros makan memori pisik ...
om ambilkan dari mesin sendiri ...
semoga menjadi manfaat dan menghilangkan keragu2an performa RC1Squid Object Cache: Version LUSCA r14850 patched: chudy r14 Start Time: Sun, 08 May 2011 04:59:10 GMT Current Time: Tue, 10 May 2011 08:38:55 GMT Connection information for squid: Number of clients accessing cache: 350 Number of HTTP requests received: 2435716 Number of ICP messages received: 0 Number of ICP messages sent: 0 Number of queued ICP replies: 0 Request failure ratio: 0.00 Average HTTP requests per minute since start: 785.8 Average ICP messages per minute since start: 0.0 Select loop called: 110827574 times, 1.678 ms avg Cache information for squid: Request Hit Ratios: 5min: 47.7%, 60min: 44.5% Byte Hit Ratios: 5min: 42.0%, 60min: 23.2% Request Memory Hit Ratios: 5min: 1.4%, 60min: 1.6% Request Disk Hit Ratios: 5min: 65.0%, 60min: 61.7% Storage Swap size: 133473482 KB Storage Mem size: 32832 KB Mean Object Size: 27.72 KB Requests given to unlinkd: 0 Median Service Times (seconds) 5 min 60 min: HTTP Requests (All): 0.06286 0.15888 Cache Misses: 0.58309 0.61549 Cache Hits: 0.00463 0.00562 Near Hits: 0.09736 0.15888 Not-Modified Replies: 0.00179 0.00179 DNS Lookups: 0.03374 0.05559 ICP Queries: 0.00000 0.00000 Resource usage for squid: UP Time: 185985.491 seconds CPU Time: 13538.534 seconds CPU Usage: 7.28% CPU Usage, 5 minute avg: 13.14% CPU Usage, 60 minute avg: 17.69% Process Data Segment Size via sbrk(): 0 KB Maximum Resident Size: 843132 KB Page faults with physical i/o: 6 Memory accounted for: Total accounted: 535717 KB memPoolAlloc calls: 430981798 memPoolFree calls: 421183822 File descriptor usage for squid: Maximum number of file descriptors: 11095 Largest file desc currently in use: 1721 Number of file desc currently in use: 840 Files queued for open: 0 Available number of file descriptors: 10255 Reserved number of file descriptors: 100 Store Disk files open: 11 IO loop method: kqueue Internal Data Structures: 4822334 StoreEntries 6863 StoreEntries with MemObjects 6788 Hot Object Cache Items 4815911 on-disk objects
TOTALS ICP : 0 Queries, 0 Hits ( 0%) HTTP: 2069808 Requests, 887340 Hits ( 43%)
Store IO Interface Stats create.calls 721538 create.select_fail 0 create.create_fail 0 create.success 721538 open.calls 1433425 open.success 1433425 open.loadav_fail 0 open.open_fail 0
-
mantap om.. :D
aku udah cubo pakai processor amd, memori 2 Gb,hdd 500 Gb. tapi pas konfigurasi cache mgmt pada lusca/squid statusnya slalu stop.
Hard disk cache system : aufs
Hard disk cache size : 30720 ( 1024 mb *30 gb) betul ngak yah ;D ???
Memory cache size : 6
Max memory object size : 4
Maximum object size : 256000
Level 1 subdirectories : 64maaf om mungkin pertanyaan udah sering, tapi masih kurang paham. kira2 di konfigurasi cache mgmt om gimana, boleh shering ngak
-
start awal di cache dir 30 gb - 50 gb saja dulu
om aja cuma 2 x 40 gb untuk ram pisik 4 gb
cache mem start awal di 16 atau 32 mb saja dulu
max memory object start dari 16 atau 32 kb
maximum object size start diangka 32 MBuntuk settingan lainnya default dari luscanya om chudy sendiri
pengalaman om, menggunakan board server lebih terasa dibanding board desktop entry
semakin besar ram pisik semakin menggigit
pattern yg terlalu agresif tidak menambah hebat kalau disk/board gak kuat nampung request I/O yg besar -
Number of clients accessing cache: 350
ini berapa user om 1000 yah
gileee :)yg sy belum dicoba di mesin produksi om msh malu ee ternyata ok jg dibanding yg RC132
-
om aku udah coba sesuai tutorial http://forum.pfsense.org/index.php/topic,34348.0.html di processor amd
tapi SWAP usage & Disk usage 0%apakah berhasil atau tidak ???bandwidth di deso aq cuman 2bm dari isp spiidy
om gambar yg di atas load balancing+squidnya ???
maaf agak sedikit cerewet jangan marah ya om ;D ;D

 -
disk object masih 800 an …
nanti terasa ketika object sudah mencapai 1 jutaan object
toh ini mesin proxy baru up khan ...dipantau saja dahulu
kalau sudah ada swap yg berlebihan, bisa dipastikan mesin proxy megap2om gak pake LB, yg diatas pake koneksi astinet 2 mbps
sudah tuntutan karena pengguna semakin berjibun dan hobinya downloader semua :Dur welcome
-
oh gitu, makasih ya om :D
-
mendekati titik klimaks dan tetap responsive …
masih survive ... ;D ;D
sengaja skrinsut diambil disaat jam sibuk ...
cache sudah max ...
tambah disk dan ram sepertinya lebih mantep lagi, semoga kesampean ...Connection information for squid: Number of clients accessing cache: 412 Number of HTTP requests received: 3806444 Number of ICP messages received: 0 Number of ICP messages sent: 0 Number of queued ICP replies: 0 Request failure ratio: 0.00 Average HTTP requests per minute since start: 887.5 Average ICP messages per minute since start: 0.0 Select loop called: 143378140 times, 1.795 ms avg Cache information for squid: Request Hit Ratios: 5min: 48.2%, 60min: 41.4% Byte Hit Ratios: 5min: 32.5%, 60min: 25.1% Request Memory Hit Ratios: 5min: 1.0%, 60min: 1.0% Request Disk Hit Ratios: 5min: 64.5%, 60min: 63.3% Storage Swap size: 147500796 KB Storage Mem size: 32724 KB Mean Object Size: 51.02 KB Requests given to unlinkd: 0 Median Service Times (seconds) 5 min 60 min: HTTP Requests (All): 0.08265 0.10281 Cache Misses: 0.55240 0.49576 Cache Hits: 0.00562 0.00678 Near Hits: 0.15888 0.11465 Not-Modified Replies: 0.00179 0.00179 DNS Lookups: 0.05559 0.05078 ICP Queries: 0.00000 0.00000 Resource usage for squid: UP Time: 257326.003 seconds CPU Time: 21571.175 seconds CPU Usage: 8.38% CPU Usage, 5 minute avg: 18.50% CPU Usage, 60 minute avg: 18.66% Process Data Segment Size via sbrk(): 0 KB Maximum Resident Size: 604932 KB Page faults with physical i/o: 21 Memory accounted for: Total accounted: 341894 KB memPoolAlloc calls: 627277319 memPoolFree calls: 621329827 File descriptor usage for squid: Maximum number of file descriptors: 11095 Largest file desc currently in use: 1497 Number of file desc currently in use: 1425 Files queued for open: 0 Available number of file descriptors: 9670 Reserved number of file descriptors: 100 Store Disk files open: 12 IO loop method: kqueue Internal Data Structures: 2897624 StoreEntries 7719 StoreEntries with MemObjects 7572 Hot Object Cache Items 2890938 on-disk objects
Store Directory Statistics: Store Entries : 2896457 Maximum Swap Size : 163840000 KB Current Store Swap Size: 147448042 KB Current Capacity : 90% used, 10% free
-
:-* top om
-
fakta berbicara …
gambar menjelaskan ...yg terakhir membuat mesin megap2 minta tolong … ;D ;D
om aja bingung, ini beneran segitu banyak atau cuma log doang atau mkn bug di report RRD nya
emang om pikirkan ... ;D yg penting tancap terus sampe klimakssemoga bermanfaat, gambar2 tsb utk referensi, "sedikiiiit" power dari pfsense
-
start awal di cache dir 30 gb - 50 gb saja dulu
om aja cuma 2 x 40 gb untuk ram pisik 4 gb
cache mem start awal di 16 atau 32 mb saja dulu
max memory object start dari 16 atau 32 kb
maximum object size start diangka 32 MBuntuk settingan lainnya default dari luscanya om chudy sendiri
pengalaman om, menggunakan board server lebih terasa dibanding board desktop entry
semakin besar ram pisik semakin menggigit
pattern yg terlalu agresif tidak menambah hebat kalau disk/board gak kuat nampung request I/O yg besarWah mau coba pake IBM SERVER yang PAKE QUAD CORE AHH siapa tau lebih Josss
-
om apa perintah untuk menampilkan pesan seperti ini??
TOTALS ICP : 0 Queries, 0 Hits ( 0%) HTTP: 2069808 Requests, 887340 Hits ( 43%)
oh iya om, kalau boleh tau paket apa saja yg om install? soalnya bandwidth om 6 mb ngak putus :D
-
itu bagian dari cachemanager …
kebetulan punya koneksi up to 16 mbps
di cut di max 6 mbps ... policy dari atasutk paket, hanya lusca + sg saja
-
om serangku ;D
bandwidth aq mungkin terlimit sama captive portal,
soalnya tiap komputer di warnet aq kasih 350kbps, pas aq cuba buka youtube dan aq ptar yg ke dua kalinya
traficnya ngak naik, lalu aq coba kasih bandwidth di cp 5000 baru naik trafic bandwidthnya… ???
aq udah baca dari forum http://forum.pfsense.org/index.php/topic,26122.0.html tapi ngak ngerti damana mau di taroh file yg mau di edit itu??
mohon bantuannya om?? :'(
/* $Id$ part of m0n0wall (http://m0n0.ch/wall) Copyrigth (C) 2009 Ermal Luçi Copyright (C) 2003-2006 Manuel Kasper <mk@neon1.net>. All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1\. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2\. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ /* pfSense_BUILDER_BINARIES: /sbin/ipfw pfSense_MODULE: captiveportal */ require_once("auth.inc"); require_once("functions.inc"); require_once("captiveportal.inc"); $errormsg = "Invalid credentials specified."; header("Expires: 0"); header("Cache-Control: no-store, no-cache, must-revalidate"); header("Cache-Control: post-check=0, pre-check=0", false); header("Pragma: no-cache"); header("Connection: close"); $orig_host = $_ENV['HTTP_HOST']; $orig_request = $_REQUEST['redirurl']; $clientip = $_SERVER['REMOTE_ADDR']; if (!$clientip) { /* not good - bail out */ log_error("Captive portal could not determine client's IP address."); $error_message = "An error occurred. Please check the system logs for more information."; portal_reply_page($redirurl, "error", $errormsg); exit; } if (isset($config['captiveportal']['httpslogin'])) $ourhostname = $config['captiveportal']['httpsname'] . ":8001"; else { $ifip = portal_ip_from_client_ip($clientip); if (!$ifip) $ourhostname = $config['system']['hostname'] . ":8000"; else $ourhostname = "{$ifip}:8000"; } if ($orig_host != $ourhostname) { /* the client thinks it's connected to the desired web server, but instead it's connected to us. Issue a redirect... */ if (isset($config['captiveportal']['httpslogin'])) header("Location: https://{$ourhostname}/index.php?redirurl=" . urlencode("http://{$orig_host}{$orig_request}")); else header("Location: http://{$ourhostname}/index.php?redirurl=" . urlencode("http://{$orig_host}{$orig_request}")); exit; } if (preg_match("/redirurl=(.*)/", $orig_request, $matches)) $redirurl = urldecode($matches[1]); if ($_POST['redirurl']) $redirurl = $_POST['redirurl']; if (!empty($config['captiveportal']['redirurl'])) $redirurl = $config['captiveportal']['redirurl']; $macfilter = !isset($config['captiveportal']['nomacfilter']); $passthrumac = isset($config['captiveportal']['passthrumacadd']); /* find MAC address for client */ $clientmac = arp_get_mac_by_ip($clientip); if (!$clientmac && ($macfilter || $passthrumac)) { /* unable to find MAC address - shouldn't happen! - bail out */ captiveportal_logportalauth("unauthenticated","noclientmac",$clientip,"ERROR"); echo "An error occurred. Please check the system logs for more information."; log_error("Captive portal could not determine client's MAC address. Disable MAC address filtering in captive portal if you do not need this functionality."); exit; } /* find out if we need RADIUS + RADIUSMAC or not */ if (file_exists("{$g['vardb_path']}/captiveportal_radius.db")) { $radius_enable = TRUE; if (isset($config['captiveportal']['radmac_enable'])) $radmac_enable = TRUE; } if ($_POST['logout_id']) { echo << <eod<br><title>Disconnecting...</title> **You have been disconnected.** EOD; captiveportal_disconnect_client($_POST['logout_id']); exit; } else if ($clientmac && $radmac_enable && portal_mac_radius($clientmac,$clientip)) { /* radius functions handle everything so we exit here since we're done */ exit; } else if (portal_consume_passthrough_credit($clientmac)) { /* allow the client through if it had a pass-through credit for its MAC */ captiveportal_logportalauth("unauthenticated",$clientmac,$clientip,"ACCEPT"); portal_allow($clientip, $clientmac, "unauthenticated"); } else if ($_POST['accept'] && $_POST['auth_voucher']) { $voucher = trim($_POST['auth_voucher']); $timecredit = voucher_auth($voucher); // $timecredit contains either a credit in minutes or an error message if ($timecredit > 0) { // voucher is valid. Remaining minutes returned // if multiple vouchers given, use the first as username $a_vouchers = split("[\t\n\r ]+",$voucher); $voucher = $a_vouchers[0]; $attr = array( 'voucher' => 1, 'session_timeout' => $timecredit*60, 'session_terminate_time' => 0); if (portal_allow($clientip, $clientmac,$voucher,null,$attr)) { // YES: user is good for $timecredit minutes. captiveportal_logportalauth($voucher,$clientmac,$clientip,"Voucher login good for $timecredit min."); } else { portal_reply_page($redirurl, "error", $config['voucher']['msgexpired'] ? $config['voucher']['msgexpired']: $errormsg); } } else if (-1 == $timecredit) { // valid but expired captiveportal_logportalauth($voucher,$clientmac,$clientip,"FAILURE","voucher expired"); portal_reply_page($redirurl, "error", $config['voucher']['msgexpired'] ? $config['voucher']['msgexpired']: $errormsg); } else { captiveportal_logportalauth($voucher,$clientmac,$clientip,"FAILURE"); portal_reply_page($redirurl, "error", $config['voucher']['msgnoaccess'] ? $config['voucher']['msgnoaccess'] : $errormsg); } } else if ($_POST['accept'] && $radius_enable) { if ($_POST['auth_user'] && $_POST['auth_pass']) { $auth_list = radius($_POST['auth_user'],$_POST['auth_pass'],$clientip,$clientmac,"USER LOGIN"); $type = "error"; if (!empty($auth_list['url_redirection'])) { $redirurl = $auth_list['url_redirection']; $type = "redir"; } if ($auth_list['auth_val'] == 1) { captiveportal_logportalauth($_POST['auth_user'],$clientmac,$clientip,"ERROR",$auth_list['error']); portal_reply_page($redirurl, $type, $auth_list['error'] ? $auth_list['error'] : $errormsg); } else if ($auth_list['auth_val'] == 3) { captiveportal_logportalauth($_POST['auth_user'],$clientmac,$clientip,"FAILURE",$auth_list['reply_message']); portal_reply_page($redirurl, $type, $auth_list['reply_message'] ? $auth_list['reply_message'] : $errormsg); } } else { captiveportal_logportalauth($_POST['auth_user'],$clientmac,$clientip,"ERROR"); portal_reply_page($redirurl, "error", $errormsg); } } else if ($_POST['accept'] && $config['captiveportal']['auth_method'] == "local") { //check against local user manager $loginok = local_backed($_POST['auth_user'], $_POST['auth_pass']); if ($loginok){ captiveportal_logportalauth($_POST['auth_user'],$clientmac,$clientip,"LOGIN"); portal_allow($clientip, $clientmac,$_POST['auth_user']); } else { captiveportal_logportalauth($_POST['auth_user'],$clientmac,$clientip,"FAILURE"); portal_reply_page($redirurl, "error", $errormsg); } } else if ($_POST['accept'] && $clientip) { captiveportal_logportalauth("unauthenticated",$clientmac,$clientip,"ACCEPT"); portal_allow($clientip, $clientmac, "unauthenticated"); } else { /* display captive portal page */ portal_reply_page($redirurl, "login",null,$clientmac,$clientip); } exit; ?></eod<br></mk@neon1.net>
-
om belum berani terapkan QoS pk CP …
kadung sudah up dan produksi
juga sudah terlalu banyak beban yg ditampung server
dari vlan, cp, otentikasi, dan beberapa package lainnya
khawatir megap2 itu server ... ;D
jadi om belum bisa memberikan pendapat lain
mungkin bisa ditanyakan langsung di thread tsb. -
kalau solusi lainnya gimana om?
ada cara lain ngak om, buat management bandwidth selain di CP?
jangan marah ya om ;D
-
masih digodok … ;)
banyak jalan menuju romautk step awal bisa dilihat thread ttg QoS yg ada ...
-
seep om,
QoS itu apa yah om??
biar aq cari jalan tikus untuk menuju ke roma.. ;D ;D
-
QoS = Quality of Service
dengan kata lain gak beda jauh dengan Traffic Shapingemang menyebalkan nyari settingan yg pas buat QoS
bukan berarti tidak bisa …
hanya belum ketemu yang pas ... :) -
semakin senang maen 64bit om kyknya ok banget lebih repson dengan lusca nya.
tos zph jg ok, server manteng trussssssss :-*