Re: Routing all sites to maintenance page

I am of course not able to come up with a solution fitting to every
single scenario out there but assuming you have different http daemons
listening at the ports 80 to 80+x or at least different pages being
available via each of these ports you could declare these ports as being
backup ports (on a per port basis) and you could bind different ports to
different services to have different error pages for different services.
Some real server ports might be backup and some other primary
ports.

A mapping like

VIP A -> primary servers 1,2,3 @ port http and backup servers 4,5,6 @
port 81

VIP B -> primary servers 2,3 @ port http and backup server 4,5,6 @
port 82

is this what you are asking for Ken?

On top of that we have introduced a way to link backup servers to primary
servers a while ago - something like:

real server A port X is using real server B port Y as backup.

R, Oliver

At 18:25 05.10.2009, Kenneth Salchow wrote:

Content-Language: en-US

Content-Type: multipart/signed;
protocol="application/x-pkcs7-signature";

micalg=SHA1;
boundary="----=_NextPart_000_1092_01CA45AE.98B27A20"

Hey­knowing what Mathew is trying to do­and looking at the configs­I had
another question for Oliver.

In the past when I’ve done this, I’ve found that a single “not available”
page for all sites becomes untenable quickly as it is too generic.
Even as an admin­it’s sometime nice to be able to see the page quickly
and know exactly which ‘site’ is down just by getting the error page,
right?

So­not knowing the ServerIron specifically, I’m assuming that for each
group of real servers, you can assign a backup server that uses a
different port? That way Mathew could, if he wanted, run a single
web server with separate “error sites” at ports 80, 81, 82, 83, etc.
? Don’t know if that’s something he’s interested in­but it might be
helpful if you gave an example??

Thanks!!

KJ (Ken) Salchow, Jr. | Manager, Technical Marketing

D 651.423.1133

M 612.868.1258

P 206.272.5555

F 206.272.5555

www.f5.com

From: [email protected]
[ mailto:[email protected] ] On Behalf Of Oliver Adam

Sent: Monday, October 05, 2009 10:36 AM

To: Load Balancing Mailing List; 'Load Balancing Mailing
List'

Subject: Re: [load balancing] Routing all sites to maintenance
page

Daiv statements are correct. You have to use so called backup servers to
do so. There are two different way to do this:

1. configure your "normal/primary" real servers together with
real servers being marked as backup servers using the backup setting
below the real server configuration:

normal real server:

server real rs101 192.168.9.101

port http

[...]

backup real server:

server real rs201 192.168.9.201

backup

port http

[...]

bind both types of servers to the virtual server and enable the feature
to activate the backup server functionality:

server virtual vs222 192.168.7.222

port http

port http lb-pri-servers

bind http rs101 http rs201 http

The backup server(s) will jump in as soon as no primary server is
available anymore or as soon as all primary servers reached their max
connection setting.

2. configure real servers and remote real servers. remote real servers do
act as backup servers per default. remote server are (by defintion) at
least one hop away from the ServerIron:

server real rs101 192.168.9.101

port http

[...]

remote real server:

server remote rs201 192.168.100.201

port http

[...]

bind both types of servers to the virtual server and enable the feature
to activate the backup server functionality:

server virtual vs222 192.168.7.222

port http

bind http rs101 http rs201 http

Both way are going to do want you are asking for. It is possible to use
backup servers per service as well - please check the documentation to
get more details about this. Per service backup servers do come in as
soon as the service they are linked to (at a given real server) is not
available anymore.

Just to add one comment:

feel free to ask questions via our official ADC community
http://community.brocade.com/adi Regards,

Oliver

At 17:09 05.10.2009, Van Ceylon, David wrote:

Content-Language: en-US

Content-Type: multipart/alternative;

boundary="_000_26D139D7C4E9784494738860A529B91E27367D4220qtomaexmbm21A_"

(long time reader and first time responder here)

I’ve dealt with this exact situation. Depending on how old your
Foundry is and what code rev, there is a way to configure your “down for
maintenance” server as a backup – even on a per-port basis if you have
9.2 code or higher. If not you can configure the server as a backup
for all your real servers as a whole. The main issue though is that
this “backup” won’t engage until all servers bound in the VIP are failing
the health-checks.

In the config, your backup server that you set up will have a “backup”
command in the real server settings. Then in your VIP, configure
the bindings for the other real servers as usual, including your backup
server and that’s about it unless you want to get more specific. If
that’s the case then post your HW/SW setup.

Good luck

David

----------

From: [email protected]
[ mailto:[email protected] mailto:[email protected] ] On Behalf Of Matthew
Schmidt

Sent: Monday, October 05, 2009 8:40 AM

To: Load Balancing Mailing List

Subject: [load balancing] Routing all sites to maintenance
page

Hi everyone. Long time reader, first time poster. We've had
an older Foundry ServerIron device sitting in front of all our sites for
some time now, but haven't set it up to do any serious load
balancing. One of the first things I would like to do is route all
requests to servers that aren't responding to a server running a page
that says we're down for maintenance.

My initial thought is that I would create "Real servers" for
each of the services currently not being managed by the load
balancer. Then, give some IPs to "virtual servers" to act
as the new frontends. From there, I'm not quite sure where to
proceed. Is there a way to tell it to go somewhere only when we
can't reach something else? Is there a better way to handle this
than with the load balancer?

Thanks,

Matt

--

Matthew Schmidt

VP of Technology, DZone, Inc.

Phone: 919.678.0300 x102

Email: [email protected]

Yahoo/AOL: mattschmidtjl

Jabber: [email protected]

_______________________________________________

lb-l mailing list

[email protected]

http://vegan.net/mailman/listinfo/lb-l

Searchable Archive:
http://vegan.net/lb/archive

http://lbdigest.com
Load Balancing Digest

http://lbwiki.com Load Balancing
Wiki

_______________________________________________

lb-l mailing list

[email protected]

http://vegan.net/mailman/listinfo/lb-l

Searchable Archive:
http://vegan.net/lb/archive

http://lbdigest.com
Load Balancing Digest

http://lbwiki.com Load
Balancing Wiki

_______________________________________________
lb-l mailing list
[email protected]
http://vegan.net/mailman/listinfo/lb-l
Searchable Archive: http://vegan.net/lb/archive
http://lbdigest.com Load Balancing Digest
http://lbwiki.com Load Balancing Wiki