|
Hello
Use the source code ipvsadm-1.26.tar.gz install the lvs tool, run the make
command, reported the following error:
[root@lvs1 ipvsadm-1.26]# make
make -C libipvs
make[1]: Entering directory `/home/software/ipvsadm-1.26/libipvs'
make[1]: Nothing to be done for `all'.
make[1]: Leaving directory `/home/software/ipvsadm-1.26/libipvs'
gcc -Wall -Wunused -Wstrict-prototypes -g -o ipvsadm ipvsadm.o config_stream.o
dynamic_array.o libipvs/libipvs.a -lnl
ipvsadm.o: In function `parse_options':
/home/software/ipvsadm-1.26/ipvsadm.c:432: undefined reference to
`poptGetContext'
/home/software/ipvsadm-1.26/ipvsadm.c:435: undefined reference to
`poptGetNextOpt'
/home/software/ipvsadm-1.26/ipvsadm.c:660: undefined reference to
`poptBadOption'
/home/software/ipvsadm-1.26/ipvsadm.c:502: undefined reference to
`poptGetNextOpt'
/home/software/ipvsadm-1.26/ipvsadm.c:667: undefined reference to `poptStrerror'
/home/software/ipvsadm-1.26/ipvsadm.c:667: undefined reference to
`poptBadOption'
/home/software/ipvsadm-1.26/ipvsadm.c:670: undefined reference to
`poptFreeContext'
/home/software/ipvsadm-1.26/ipvsadm.c:677: undefined reference to `poptGetArg'
/home/software/ipvsadm-1.26/ipvsadm.c:678: undefined reference to `poptGetArg'
/home/software/ipvsadm-1.26/ipvsadm.c:679: undefined reference to `poptGetArg'
/home/software/ipvsadm-1.26/ipvsadm.c:690: undefined reference to `poptGetArg'
/home/software/ipvsadm-1.26/ipvsadm.c:693: undefined reference to
`poptFreeContext'
collect2: ld 返回 1
make: *** [ipvsadm] 错误 1
OS Version: Redhat enterprise 6;
Kernel Version: 2.6.32-71.el6.x86_64;
祝工作顺利！！！
Name：饶俊明
Tel：0755-86681223/15989549894
Mail：raojm@xxxxxxxxx
Web：www.mibuo.com
Comp：深圳饭否
2012-02-10
发件人： lvs-users-request
发送时间： 2012-02-09 20:01:36
收件人： lvs-users
抄送：
主题： lvs-users Digest, Vol 109, Issue 6
Send lvs-users mailing list submissions to
lvs-users@xxxxxxxxxxxxxxxxxxxxxx
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.graemef.net/mailman/listinfo/lvs-users
or, via email, send a message with subject or body 'help' to
lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
You can reach the person managing the list at
lvs-users-owner@xxxxxxxxxxxxxxxxxxxxxx
When replying, please edit your Subject line so it is more specific
than "Re: Contents of lvs-users digest..."
Today's Topics:
1. -x (or --u-threshold) Limits Not Working Right (Robinson, Eric)
2. Persistence When CIP Loses Network (Jason Ledford)
3. Re: -x (or --u-threshold) Limits Not Working Right
(Julian Anastasov)
4. Re: Announcement: Piranha being retired in Fedora 17
(L.S. Keijser)
------------------------------------------------------------
From: "Robinson, Eric" <eric.robinson@xxxxxxxxx>
To: <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: [lvs-users] -x (or --u-threshold) Limits Not Working Right
Date: Wed8 Feb 2012 08:35:37 -0800
I have a pair of Windows terminal servers behind my load balancer. I
want to make sure that neither terminal ever gets more than 150
connections, so I specified the -x (or --u-threshold) parameter
accordingly. After a few hours, some users started complaining that they
could not connect, and when I ran the ipvsadm command to check the
number of connections that had been established, both servers were at
around 120. I removed the -x parameter and the users could then connect.
Why the difference between the threshold number that we set and the
actual maximum number of realserver redirections that LVS allows?
--
Eric Robinson
Disclaimer - February 8, 2012
This email and any files transmitted with it are confidential and intended
solely for lvs-users@xxxxxxxxxxxxxxxxxxxxxxx If you are not the named addressee
you should not disseminate, distribute, copy or alter this email. Any views or
opinions presented in this email are solely those of the author and might not
represent those of Physicians' Managed Care or Physician Select Management.
Warning: Although Physicians' Managed Care or Physician Select Management has
taken reasonable precautions to ensure no viruses are present in this email,
the company cannot accept responsibility for any loss or damage arising from
the use of this email or attachments.
This disclaimer was added by Policy Patrol: http://www.policypatrol.com/
------------------------------------------------------------
------------------------------------------------------------
From: Jason Ledford <jledford@xxxxxxxxxxxx>
To: "lvs-users@xxxxxxxxxxxxxxxxxxxxxx" <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: [lvs-users] Persistence When CIP Loses Network
Date: Wed8 Feb 2012 21:41:42 -0500
I must have something setup wrong. My goal is to load balance microsoft
terminal servers and have clients connect back to the same server within the
timeout period. This works fine if you log out of the terminal server or x out
of the rdp session, when you reconnect you are directed to the same server, and
if you session was left open you get the same session. Apparently thats where
my testing stopped. If you drop the network on the client and reconnect to the
terminal session, I would expect you to return to the same server, but you
don't. The client gets the same ip and their hostname hasn't changed. What is
different about logging out or X out vs losing network? below is my
ldirectord.cf file. Let me know what other information you need to offer some
advice. Beyond this it has been working like a champ for ~2 years and someone
recently brough this to my attention. Thanks for looking.
#global options
checktimeout=20
checkinterval=4
autoreload=yes
quiescent=no
virtual = 10.72.5.4:3389
protocol=tcp
scheduler=wlc
checkport=3389
checktype=connect
real=10.72.5.5:3389 gate 1
real=10.72.5.6:3389 gate 1
real=10.72.5.7:3389 gate 1
# real=10.12.1.40:3389 gate 1
persistent=10800
------------------------------------------------------------
------------------------------------------------------------
From: Julian Anastasov <ja@xxxxxx>
To: "Robinson, Eric" <eric.robinson@xxxxxxxxx>
Subject: Re: [lvs-users] -x (or --u-threshold) Limits Not Working Right
Date: Thu9 Feb 2012 10:03:11 +0200 (EET)
Hello,
On Wed, 8 Feb 2012, Robinson, Eric wrote:
> I have a pair of Windows terminal servers behind my load balancer. I
> want to make sure that neither terminal ever gets more than 150
> connections, so I specified the -x (or --u-threshold) parameter
> accordingly. After a few hours, some users started complaining that they
> could not connect, and when I ran the ipvsadm command to check the
> number of connections that had been established, both servers were at
> around 120. I removed the -x parameter and the users could then connect.
>
> Why the difference between the threshold number that we set and the
> actual maximum number of realserver redirections that LVS allows?
The algorithm counts both active and inactive connections
when checking the threshold, i.e. states before and after ESTABLISHED.
It hurts that SYN states are counted as inactive because it causes bad
scheduling even for WLC/LC schedulers. Do you see that active+inactive
sum exceeds the threshold of 150?
Regards
--
Julian Anastasov <ja@xxxxxx>
------------------------------------------------------------
------------------------------------------------------------
From: "L.S. Keijser" <leon@xxxxxxxx>
To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: [lvs-users] Announcement: Piranha being retired in Fedora 17
Date: Thu09 Feb 2012 11:07:37 +0100
On Fri, 2012-01-27 at 13:26 +0000, Michael Schwartzkopff wrote:
> > How do you keep your ldirector config files in sync? Shared filesystem,
> > svn/cvs, or just rsync/scp?
> >
> > I've got pretty much the same config - A cloned ldirectord resource on
> > pacemaker. I'm using DRBD+GFS2, but it seems overkill. Just curious what
> > others are doing for similar implementations.
> >
> > David
>
> Yes. DRBD+GFS is overkill. Since most configs change only per year I do the
> sync manually. rsync in a cronjob would be the automatic alternative.
>
A bit late to reply, but anyway ..
ldirectord has an option execute a command as soon as the config
changes. For example:
autoreload = yes
callback = "/usr/local/bin/sync_ldirectord"
Then the contents of sync_ldirectord could be something like this:
rsync -aq /etc/ldirectord.cf lb02::lvs
or scp , or whatever you want. Then as soon as you change the
ldirectord.cf on one node, it is synced to the other. This is how I've
been doing it for years. Usually rsyncd on the other side so I don't
need to allow passwordless ssh or mess with scponly.
regards,
Léon
------------------------------------------------------------
_______________________________________________
lvs-users mailing list
lvs-users@xxxxxxxxxxxxxxxxxxxxxx
http://lists.graemef.net/mailman/listinfo/lvs-users
_______________________________________________
Please read the documentation before posting - it's available at:
http://www.linuxvirtualserver.org/
LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to http://lists.graemef.net/mailman/listinfo/lvs-users
|
