Discuss

discuss@lists.centos.org

34758 discussions

IP forwarding and OpenVPN
by Bart Schaefer 05 Nov '10

05 Nov '10

I have a CentOS 4 server that acts as a gateway for a small LAN. The lower half of the 192.168.144 address space is the LAN on eth1, the upper half is the WAN on eth0, and the default route is to 192.168.144.254 which is my DSL router; this has been working fine for years. However, it's recently become convenient to connect the server to a VPN from time to time, for which I've set up OpenVPN. This works for connections originating from the server itself, but breaks for machines on the LAN when accessing IPs in the ranges that are routed to the VPN. Connections to IPs not in the private network still work as before. Can anyone advise what I may need to change to configure the server to forward packets to the VPN? Pointers to documentation are welcome. Thanks. /etc/sysctl.conf has: net.ipv4.ip_forward = 1 net.ipv4.conf.default.rp_filter = 1 net.ipv4.conf.default.accept_source_route = 0

3 12

new apache mod on centos? Google page speed?
by Bob Hoffman 05 Nov '10

05 Nov '10

Just saw this article http://news.yahoo.com/s/mashable/20101103/tc_mashable/google_helps_autotune… (/rant - stupid seo links, lame /rant off) it mentions it is set up for linux distros, notably centos. Anyone use this yet? Affect the auto yum update at all?

2 1

Re: [CentOS] was, PATA Hard Drive woes, is "SMART"
by m.roth＠5-cent.us 04 Nov '10

04 Nov '10

Folks, Thanks a *lot* to those who suggested I try fsck -cc - that took six hours, and at the end, I was standing there, and it tells me that there's two back blocks, one was /dead/base.pp, and... (wait for it) bad blocks, in node 1. I said yes, fix them, and rebooted, and I don't see complaints. mark "and there was much rejoicing"

3 2

orphan inodes deleted issue
by David S. 04 Nov '10

04 Nov '10

Dear All, My servers running on CentOS 5.5 x86_64 with kernel 2.6.18.194.17.4.el gigabyte motherboard and 2 harddisks (seagate 500GB). My CentOS box configured RAID 1, yesterday and today I had the same problem on 2 servers with same configuration. See the following error messages for details: EXT3-fs: INFO: recovery required on readonly filesystem. EXT3-fs: write access will be enabled during recovery. kjournald starting. Commit interval 5 seconds EXT3-fs: md1: orphan cleanup on readonly fs ext3_orphan_cleanup: deleting unreferenced inode 2010087 ext3_orphan_cleanup: deleting unreferenced inode 259361 ext3_orphan_cleanup: deleting unreferenced inode 972510 ext3_orphan_cleanup: deleting unreferenced inode 2009841 ext3_orphan_cleanup: deleting unreferenced inode 2010102 ext3_orphan_cleanup: deleting unreferenced inode 2010079 ext3_orphan_cleanup: deleting unreferenced inode 2010078 ext3_orphan_cleanup: deleting unreferenced inode 2009831 ext3_orphan_cleanup: deleting unreferenced inode 2010001 ext3_orphan_cleanup: deleting unreferenced inode 2009991 ext3_orphan_cleanup: deleting unreferenced inode 2009989 ext3_orphan_cleanup: deleting unreferenced inode 2009811 ext3_orphan_cleanup: deleting unreferenced inode 2009806 ext3_orphan_cleanup: deleting unreferenced inode 2009978 ext3_orphan_cleanup: deleting unreferenced inode 2009976 ext3_orphan_cleanup: deleting unreferenced inode 2009972 ext3_orphan_cleanup: deleting unreferenced inode 2009828 ext3_orphan_cleanup: deleting unreferenced inode 2009964 ext3_orphan_cleanup: deleting unreferenced inode 2009865 ext3_orphan_cleanup: deleting unreferenced inode 2009863 ext3_orphan_cleanup: deleting unreferenced inode 2009830 ext3_orphan_cleanup: deleting unreferenced inode 2009809 ext3_orphan_cleanup: deleting unreferenced inode 2009823 ext3_orphan_cleanup: deleting unreferenced inode 2009856 ext3_orphan_cleanup: deleting unreferenced inode 2009807 ext3_orphan_cleanup: deleting unreferenced inode 2009799 ext3_orphan_cleanup: deleting unreferenced inode 2009794 EXT3-fs: md1: 27 orphan inodes deleted EXT3-fs: recovery complete. It's my array: $ df -h Filesystem Size Used Avail Use% Mounted on /dev/md1 9.5G 735M 8.3G 8% / /dev/md7 38G 6.7G 30G 19% /var /dev/md6 15G 4.5G 9.1G 33% /usr /dev/md5 103G 45G 54G 46% /backup /dev/md3 284G 42G 228G 16% /home /dev/md2 2.0G 214M 1.7G 12% /tmp /dev/md0 243M 24M 207M 11% /boot I've been searching on google but I can't find the explanation for this problem. It's a BUG? :D Thank You very much :D -- - -- Best regards, David http://blog.pnyet.web.id

4 5

generate signing subkey
by Tim Dunphy 04 Nov '10

04 Nov '10

hello all I need to generate a subkey in gpg in order to sign my emails. I already have a main key. how do I generate the subkey I need to encrypt emails? I have already run gpg --gen-key thanks -- Here's my RSA Public key: gpg --keyserver pgp.mit.edu --recv-keys 5A4873A9 Share and enjoy!!

2 2

failure to import ldif into ldap
by Tim Dunphy 04 Nov '10

04 Nov '10

Thanks all.. I have read the man of ldif.... your advice has gotten me quite far both in my current implementation and in my overall understanding of LDAP which I am hoping grows with each passing day. In my attempt to build my current directory, I have taken a dump of my last successful implementation (which was created on FreeBSD 8.1) and substituted values for the dc=company and dc=com values with the correct ones for the current directory (attempting to implement under CentOS 5.4) and even tho the correct schemas are in place it is choking on this entry: # defaults, sudoers, Services, acadaca.com dn: cn=defaults,ou=sudoers,ou=Services,dc=acadaca,dc=net objectClass: top objectClass: sudoRole cn: defaults description: Default sudoOption's go here And again I should have all the schemas in place to make this work... include /etc/openldap/schema/core.schema include /etc/openldap/schema/cosine.schema include /etc/openldap/schema/inetorgperson.schema include /etc/openldap/schema/nis.schema include /etc/openldap/schema/misc.schema inlcude /etc/openldap/schema/sudoers.schema include /etc/openldap/schema/openldap.schema [root@ldap ldif]# ldapadd -h ldap -a -w secret -x -D "cn=Manager,dc=acadaca,dc=net" -f /home/tim/txt/ldif/acadaca-master.ldif adding new entry "cn=defaults,ou=sudoers,ou=Services,dc=acadaca,dc=net" ldapadd: Invalid syntax (21) additional info: objectClass: value #1 invalid per syntax Why this ldif will work in one directory but not another is a mystery at this point.. thanks again -- Here's my RSA Public key: gpg --keyserver pgp.mit.edu --recv-keys 5A4873A9 Share and enjoy!!

1 0

Copying root partition
by Timothy Murphy 04 Nov '10

04 Nov '10

I recently transferred the / partition on my CentOS server from a small disk to a large disk, using "rsync -auvz". This works fine, except that I get dozens of selinux warnings when I re-boot. I'm running selinux in permissive mode. Is there any way to make sure that all the files in a partition are kosher as far as selinux is concerned? -- Timothy Murphy e-mail: gayleard /at/ eircom.net tel: +353-86-2336090, +353-1-2842366 s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland

5 4

Web statistics - w3perl
by Jussi Hirvi 04 Nov '10

04 Nov '10

I am considering a richer alternative to good old Webalizer, in a webhotel (multidomain) setting. Any experience with w3perl? The demo at least looks impressive. http://www.w3perl.com - Jussi -- Jussi Hirvi * Green Spot Topeliuksenkatu 15 C * 00250 Helsinki * Finland Tel. +358 9 493 981 * Mobile +358 40 771 2098 (only sms) jussi.hirvi(a)greenspot.fi * http://www.greenspot.fi

5 5

Re: [CentOS] PATA Hard Drive woes
by Lamar Owen 04 Nov '10

04 Nov '10

>From: Keith Roberts <keith(a)karsites.net> >On Wed, 3 Nov 2010, Lamar Owen wrote: >> Might want to check the power supply as well. Bad/flakey >> power can indeed case damage to the drive surface; been >> there, done that, have two Maxtor 250GB drives with >> scribbled servo data to prove it. >OK. > I'm running the server from an APC UPS Back-UPS 650, so > there should not be any glitches in the power supply, should > there? Probably not on the AC side, although the Back-UPS 650 isn't a full online UPS but a switching standby UPS (full online, like the APC Symmetra 16KVA units I have here) rectify to DC, float the batteries at all times, and run the output from inverter all of the time (unless they're switched to bypass). The SmartUPS 1400RM I had in front of the PC that suffered the glitchy power is, unless I'm mistaken, also a full online pure sinewave UPS like the Symmetra, and is still in service (I checked its output on my oscilloscope first, though). No, I was referring to the output DC voltages (+12V, +5V, +3.3V,-5V, and -12V) from the power supply inside the system. In addition to my own personal RAID1 of 250GB drives, I also, a different time, lost a RAID5 array of 15K 36GB SCSI drives in a Dell 1600SC server; testing the power supply showed lots of noise and complete dropouts of a few milliseconds duration on the drive connectors' 5V supply pins. Completely and thoroughly scrambled the servo data on the Hitachi drives. Meaning they didn't just start showing bad sectors; they started getting seek errors. The 5V line on the drive connectors was reading an AC RMS of 4V superimposed on the +5V, yielding an effective DC voltage of 4V. Happened over a period of three weeks, during which time I had a number of mysterious failures (the Hitachi drives were error-correcting so well that by the time they started reporting errors, it was way past too late, and it became impossible for the Hitachi drives to even power up). I found that the power supply in question, upon investigation, provided the motherboard (where the DC power sensors on that box are) with clean 5V, and the drives were powered from a separate 5V rail, meaning the Dell management system wasn't seeing the power problems. A simple power supply tester with a built-in meter can be bought for less than $20; a more thorough power analyzer will run more than that. But even the simple one caught the failing Dell 1600SC supply. It took an oscilloscope to test the Antec in my personal box; turned out it was a cold solder joint in the Antec. A new power supply is less expensive than the equivalent labor it took to fix the Antec. I keep a known good 500W ATX 12V server-grade (8 pin 12V plug with adapters, and 24-pin ATX plug with 20-pin adapter) around for testing; that's one of the very first things I check when a PC is brought in that is flaky. (The very first thing is the dust accumulation, and the second thing is the heatsink compound). One of the first things I do on any CentOS system I put together is install lm_sensors and gkrellm (gkrellm from a third-party repo). I then enable all the motherboard sensors that are available in the gkrellm plugins, and run it (either local GUI or through ssh X forwarding to my central monitoring PC). On supermicro boards I install SuperODoctor for Linux, available on the supermicro site. The GUI runs well (there are some odd dependencies, however) and will e-mail you on alarm conditions that you can set. These include fan RPM, temperatures, and voltages. The CLI program isn't quite so sophisticated, but it can be run periodically and the result sent by e-mail for health checks. Drives that are having trouble will show up with high iowaits; run iostat (from the sysstat package) and look at the await result. Long awaits mean the drive is having trouble (or it has firmware issues like WD's EARS and EADS drives have in RAID configurations).

1 0

kernel bug fixed in later kernels
by Philip Manuel 04 Nov '10

04 Nov '10

Does anyone know if this is going to be fixed in CentOS ? // test case code below // expected behaviour is "all good" and a return code of zero. // this works on 2.6.35-1 // does not work on 2.6.18-194.17.4.el5 from Centos 5.5 // think this bug was fixed in git commit 8948e11f450e6189a79e47d6051c3d5a0b98e3f3 // around 2.6.22 // http://www.mail-archive.com/git-commits-head@vger.kernel.org/msg11654.html #include <fcntl.h> #include <stdio.h> #include <sys/stat.h> #include <sys/types.h> #include <unistd.h> int main(int argc, char** argv) { int fd = open("foo", O_RDONLY); if (setuid(1000)) { printf("could not setuid, run as root with correct uid\n"); return 1; } char proc_name[1024]; sprintf(proc_name, "/proc/self/fd/%d", fd); struct stat stat_buf; int rc = stat(proc_name, &stat_buf); if (rc == 0) { printf("all good\n"); } else { printf("busted, could not access %s\n", proc_name); } return rc; }

2 2

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss