Difficulty with Virtual Named hosts on Web Ser

I am having a difficult time with creating Several Virtual Named Hosts with Apache2. I am trying not to use the Apache Manual modifications to the config file, sothat I can stil manage cia Server Admin. I am running OSX 10.5.2 and have several domains registered . Additionally I have a DNS Server running on the Xserver the primary zone is yadayadayada.com with zone as follows;
yadayadayada.com. Primary Zone
www.yadayadayada.com. ALIAS yadayada.com.
machineyada Machine 74.11.11.11
HomePage Service machineyada.yadayadayada.com:80
I have reverse Zones, and a Secondary zones setup for my DSL Service provider and my Domain host. I have two DSL Lines setup using a Cisco router , the cisco router provides DHCP , and multiplexes the two DSL lines to share bandwidth.
the web is setup as follows;
Domain_Name Address Port
* 80
yadayadayada.com. * 80
yadayadayada1.com * 80
I dont have Realms setup and all other Web config parameters are set as default
my name servers are registered , and have propagated verified by a nslookup. I have them setup to point to the same phydical server using the multiplexed ip adresses i.e.
ns1.yadayadayada.com 74.11.11.11
ns2.yadayadayada.com 74.11.11.12
when I browse to either www.yadayadayada.com or www.yadayadayada1.com I get the default website, when the default web is deactivated i receive server not available .
According to the documentation once you have a unique domain name and a Zone for that domain you should be able to host multiple Named Virtual hosts, however after explicitly following the "Vague" directions I am still not able to get the configuration to return the correct sites. does anyone have any experiance with Virtual Named hosts, or must i resort to the Apache Docs ?
Thanks for the assistance in advance.

Obfuscating the domain name often leads to confusion - for example are errors introduced via typos in the obfuscation?
Given the information that you've posted, your DNS isn't setup correctly.
You're trying to access 'www.yadaydayada.com', but you don't have any 'www' record in your 'yadayadayada.com' zone file. However, since you seem to get a response from the web server clearly you have some kind of additional DNS setup that you're not posting here.
In any case, as far as Apache is concerned, there is no relationship between 'yadayadayada.com' and 'www.yadayadayada.com'. They are two completely different hostnames and I wouldn't expect Apache to serve yadayadayada.com's web site when you're trying to access 'www.yadayadayada.com'. You'll need to either assign a server alias for the other domain, or add an additional Virtual host for 'www.yadayadayada.com' (and make sure the DNS points to the right place, of course).

Similar Messages

Virtual Named Hosts

I am having a difficult time with creating Several Virtual Named Hosts with Apache2. I am trying not to use the Apache Manual modifications to the config file, sothat I can stil manage cia Server Admin. I am running OSX 10.5.2 and have several domains registered . Additionally I have a DNS Server running on the Xserver the primary zone is yadayadayada.com with zone as follows;
yadayadayada.com. Primary Zone
www.yadayadayada.com. ALIAS yadayada.com.
machineyada Machine 74.11.11.11
HomePage Service machineyada.yadayadayada.com:80
I have reverse Zones, and a Secondary zones setup for my DSL Service provider and my Domain host. I have two DSL Lines setup using a Cisco router , the cisco router provides DHCP , and multiplexes the two DSL lines to share bandwidth.
the web is setup as follows;
Domain_Name Address Port
* 80
yadayadayada.com. * 80
yadayadayada1.com * 80
I dont have Realms setup and all other Web config parameters are set as default
my name servers are registered , and have propagated verified by a nslookup. I have them setup to point to the same phydical server using the multiplexed ip adresses i.e.
ns1.yadayadayada.com 74.11.11.11
ns2.yadayadayada.com 74.11.11.12
when I browse to either www.yadayadayada.com or www.yadayadayada1.com I get the default website, when the default web is deactivated i receive server not available .
According to the documentation once you have a unique domain name and a Zone for that domain you should be able to host multiple Named Virtual hosts, however after explicitly following the "Vague" directions I am still not able to get the configuration to return the correct sites. does anyone have any experiance with Virtual Named hosts, or must i resort to the Apache Docs ?
Thanks for the assistance in advance.

Obfuscating the domain name often leads to confusion - for example are errors introduced via typos in the obfuscation?
Given the information that you've posted, your DNS isn't setup correctly.
You're trying to access 'www.yadaydayada.com', but you don't have any 'www' record in your 'yadayadayada.com' zone file. However, since you seem to get a response from the web server clearly you have some kind of additional DNS setup that you're not posting here.
In any case, as far as Apache is concerned, there is no relationship between 'yadayadayada.com' and 'www.yadayadayada.com'. They are two completely different hostnames and I wouldn't expect Apache to serve yadayadayada.com's web site when you're trying to access 'www.yadayadayada.com'. You'll need to either assign a server alias for the other domain, or add an additional Virtual host for 'www.yadayadayada.com' (and make sure the DNS points to the right place, of course).

Need Help with Virtual Host

a quick Explaination of what I'm trying to do.
We have a DNS name, "our_intranet.com". This DNS points to a load balancer which routes the call to "our_weblogic_server:7015". When a request with this DNS comes in, we really want to go to our_weblogic_server:7015/portal/site/our_intranet. Can this be acheived using a Virtual Host? And if so, could give an example of creating a virtual host?
Thanks,
Tom.

Thank you James for your reply,
I do appreciate you taking a moment to respond and include some suggestions. At this point I feel 90% certain that the answer lies within WebLogic configuration somewhere.
Discoverer is a J2EE application as I understand it and based on that belief it is probably using the standard J2EE servlet API calls such as ServletRequest. getScheme() (the protocol portion of the URL such as http, https, ftp), ServletRequest.getHostname(), and ServletRequest.getPort(). Those API calls return values derived by the server either from the connection itself or from server configuration.
The URL's being returned by the web application are not static either. Although the protocol and the port never change (always http://...:80/) the server hostname does. If I add an entry to my host file for fred.somewhere.com and then hit the server using that server name then the URL's returned by the web application are http://fred.somewhere.com:80/, and if I use barney.somewhere.com I get a similar result. Now, the web application could be reading the "Host" HTTP header out of the request itself to derive that information but why go through all that trouble when the Servlet API gives you a standard way to do it.
For this reason I believe Discoverer (the web app causing our grief) is getting this information from WebLogic even though I cannot verify that with absolute certainty. I wish someone from the Disco area could verify that. I have had a support request open with Oracle for over a week but so far they haven't been able to tell me where they get that information from (which I'm sure is just because I'm not speaking with a developer). I may actually in fact pass on this question to the Discoverer forums you suggested. If someone can at least tell me if Discoverer is getting this information from the application server then I can at least pass that along as confirmation to WebLogic folks.
I have set up an Apache web server as a reverse proxy server for http and https in front of this particular server as an experiment and get similar results with the server name. Whether I hit Apache via http or https I always get back http://servername:80/ from the back-end WLS. The fact is WebLogic isn't even running on port 80 which makes it even more interesting why the URL's we get back should point at 80 instead of the port WLS is listenning on. Since there is no port or protocol information contained within the HTTP request sent by client browsers (only the Host header which just has the hostname) that may have something to do with why the server doesn't know how to provide correct values, but that doesn't explain why it would choose to tell the web application port 80 instead of the port it is actually listening on.
As far as tracking down how to configure this beast it might be complicated by the fact that I don't think this is just WebLogic Server. This is Fusion Middleware. I didn't think that included Apache HTTP Server, but actually I can see there is a directory for Apache on the server identified as ohs (stands for Oracle HTTP Server). Some of the help in Enterprise Manager (or maybe it was in the WebLogic administration web console) actually talks about Oracle HTTP Server which they say is based on Apache HTTP Server and in fact the configuration files are consistent with Apache. That doesn't mean I don't believe WebLogic folks like yourself that say Apache is not a part of WebLogic server (which seems like a pure Java application) but maybe it is a part of Oracle Fusion Middleware. They might just have it there for other stuff working together with WLS (such as Oracle Web Cache) or for load balancing as you mentioned. If it is there then it might either complicate things, or might provide a possible solution. If it really is running there in front of WLS and we can use some Apache configuration values to override URL information that the server passes on to the web applications maybe it could help. I did see some processes running on the same server with a name something like HTTPD.worker. My Apps DBA's (who are also learning as they go with WLS) said they did zero Apache configuration as part of the install however and I know zero about how (or if) Apache is working together with WLS.
I tried going down the path of setting up a custom Network Channel and Virtual Host because a) it seemed like a logical place to look and b) I had seen some references to them on the web but as I tried to step through the configuration I admit I never saw any settings that related to protocol overrides. I have seen host and port overrides somewhere along the way but never a protocol override.
I'm confused that this seems so complicated. I mean, aren't there a lot of people out there running their servers behind a hardware load balancer with SSL termination at the load balancer (it handles the https decryption and then forwards the requests unencrypted to the back-end servers)?
Your link to the "Corecoreserverserverprotocolshttptitle.html" looks like just the right place (almost). If it only had Frontend Port and Frontend Protocol as well as Frontend Host (and instead of Frontend HTTP Port and Frontend HTTPS Port). So frustrating to be so close but miss. I do appreciate the pointer though.
Funny that you should bring up the post on WebLogic Wonders because that is actually the example I tried to follow when I attempted to set up a network channel and virtual host. I aborted before the deploy because the deploy wasn't working quite how I expected and I had little confidence it was going to help me at that point (because I hadn't been presented with any setting that allowed me to override the protocol (http/https) value for a front-end system. I'm still not quite sure how the Network Channel works together with the server/domain/virtual host so perhaps I need to do a little more reading.
If anyone understands exactly what I am trying to achieve though and can give me some direction it would be much appreciated. It seems straight forward to me:
We have an F5 hardware load-balancer that runs here:
https://external-servername:443/....
It uses hardware SSL accelleration to decrypt the incoming request and if it sees /discoverer/ in the URL it then forwards the request unencrypted to a back-end WLS here:
http://internal-servername:8888/...
We would be happy if any connection that server recieved it always overrides the scheme (https instead of http), server (external instead of internal...it actually does this properly already), and port (443 instead of 80). In a perfect world it might behave differently if we connect internally directly to the server (in other words if we don't access it through the load balancer then it provides URL information to continue accessing it directly) but that is just icing on the cake which probably would require virtual host configuration to differentiate the two. We'd be ecstatic if we could simply override those values all the time for the entire server to some constant values that we set.
Thanks,
Eric R.

ACE probing of "virtual hosts" on web servers

We got some web servers that hosts multiple web sites and would like to probe each of these sites. Is there a way to configure a probe to present the request for a particular web site versus the other without falling into the catch all clause of the http server configured on the box ?
Thanks

Hello Folks,
I wonder if you could get away with just using the built-in HTTP probe rather than the scripted probe? With the built-in HTTP probe, you can configure all sorts of HTTP headers, including the Host header which might be all you need for your server hosting multiple sites.
Just to clarify on the support for scripted probes that TAC supports: If the script is downloaded from cisco.com, and is not modified, then TAC will support it. It is not supported if it is modified, and TAC will not support custom-created scripts.
Hope this helps,
Sean

Solaris as a guest OS with Virtual PC running Mac OS X as a host OS

Trying to instal Solaris as a guest OS - 5 times with Connectix Virtual PC 6 (Mac OS X - host OS) and once with MS Virtual PC 7. No success !
Everything's going well (Solaris detected video card and monitor) text based instal and GUI-based(finished instal CD1), until you need to reboot, to continue to proceed next.
After reboot message: X-window X11 cannot to be found but Install Lancher(GUI) is starting and ... unspectualy crushed.
Sort of message press return to reboot. rebooting.
loging with text:
root and password.
The question: How to re-start Install Lancher?
xinit : not found
startx: not found
Trying with VESA video card with minimum video effect - ZERO NOTHING !!!
THE QUESTION: HOW TO RE-START INSTALL LANCHER ? HELP will be appreciated
P.S. I have Mac Mini with Mac OS X Tiger There is no problem to install any Linux OS with Virtual PC 6 or even VP 7
I think Sun have to do something to make Solaris x86 as a guest OS run on
Virtual PC

ph0enix2002:
sys-unconfig [ENTER]
select XOrg when and entere other various config parameters when prompted
Please be more specific What is "entere other various config parameters when prompted" ?
What is your host system ? Mac OS X or MS Windows?
What is your version Virtual PC ?
How to re-start install lancher? I stuck whith first CD installed and must continue with second install CD 2 and all other stuff
[img] http://www.pix05.com/f/solaris_picture_12997.gif
How to re-start install lancher ?

Wf component container not starting with Virtual host

Hi,
We have setup fresh instance of R 12.1.1 on Solaris with Virtual host. Did installation with rapidwiz -servername virtual.
Now when we end setting up wfmailer setup it refused to start saying component services are not up. While looking at it, it is still looking for Physical host rather Virtual. Can anyone suggest how to resolve this.
Thanks

Post in 11i thread closing it

Configuring webclient with Virtual Host

i configure portal with a virtual host.
Portal it is ok. i can navigate and login with the virtual host server. But the collaboration portlets (Message portlet, calendar portlet) link to the original server.
what can i do to work also with virtual host?
my original server is the internal host name.
the virtual host is the public server name.
thanks in advance
Regards
giorgio

no problem from flex side.. need to escape crossdomain.xml
file in htacess for codeigniter i.e RewriteCond $1
!^(index\.php|images|application|uploads|robots\.txt|crossdomain\.xml)
in htaccess file.

Getting web host name, web-port, oral-report server name and virtual path

Hi.
I have been using oralce forms 6i on oracle application server 10g. I have three servers installed on remote locations.I just want that when i am hosting an application on a server, the web host name, web port , oracle report server name and the path of directory should be selected in a dynaminc variable at application level at run time. In this manner, I will be able to make a single application for all three servers. Any other solution to this problem is also welcomed. Please help anyone and oblige me.

Hi Rajesh,
Actually Web AS Host name is <Servername>.<domain name>:<Port no>
Web AS Path for R/3 is </sap/bc/bsp/sapd>
Refer the following link.
<a href="https://forums.sdn.sap.com/thread.jspa?threadID=402561">https://forums.sdn.sap.com/thread.jspa?threadID=402561</a>
<a href="https://www.sdn.sap.com/irj/sdn/thread?threadID=265076">https://www.sdn.sap.com/irj/sdn/thread?threadID=265076</a>
Regards
Hemalatha
/Assign points if helpful/

10.9 mail with virtual host

Anyone have experience in setting up mail server on OSX 10.9 with virtual domain (postfix to serve more than one domain)? Coz the old guide (https://support.apple.com/kb/HT3878) with workgroup manager should no longer apply, right?

What I would like to do is to setup the following:-
1. setup a MX Server for two domains (domaina.com and domainb.com)
2. setup some mailbox whereas some of them will be stored locally while some other will be forwarded to other server (e.g. [email protected] would be stored locally whereas [email protected] would be forwarded to [email protected] on some other server sitting in internal network)
3. setup POP3 access so that mail client can connect to this server to retrieve the mail remotely.
4. All email account should not have direct remote access to this mac mini server.
How can I achieve this?

Set up reverse DNS for virtual mail hosting

I need a bit of server configuation advice.
I have a static IP and two public domains on a Snow Leopard server connected using NAT behind a firewall - with the necessary port forwarding to ensure all works.
1. abc.com is my primary domain on the server - server.abc.com
2. I have xyz.com set up as a virtual domain and also as a virtual mail host
This setup has worked well for a long time but I have found that emails to [email protected] are going missing. If I check my mx records using one of the web based tools it show an error on the reverse dns for server.xyz.com showing a reverse DNS of server.abc.com.
So the question - is it possible to have secondary 'virtual' DNS record on the server so reverse DNS works for the virtual mail host xyz.com? If not how do I handle the reverse DNS problem which i think is causing some external mail server to reject mail due to the inconsistency on the reverse DNS lookup?
Many thanks for any suggestions

SMTP requires a DNS A record.
A DNS A record is also known as a machine record.
A DNS A record inherently means that forward DNS and reverse DNS will match.
The forward translation translates the host name to the IP address.
The reverse translation translates the IP address to host name.
When the full translation produces the same host name, that's an A record.
DNS CNAME records are aliases, and are used for virtual hosts.
CNAME records inherently do not match the reverse DNS translations.
To get your configuration to work, your server must have an A record.
That means forward and reverse DNS will match.
Any of the virtual hosts within your mail server then all use an MX pointing at the A record host.
If you have your DNS hosted somewhere other than your ISP, then you'll need your ISP to set up a DNS PTR.
The DNS PTR is the reverse translation; address to name.
If you have your own DNS services within your network (as would be typical with a privately-addressed NAT'd network), set that up as a virtual host within SMTP.
Here is some related reading on external (public) DNS, as related to SMTP servers and such.

Hosting 2 web servers

Is it possible for me to host 2 web servers using 1 public ip, if I do a nat statement to both of them, how will it know which server to use from the outside ?

Hello Carl,
I can see that you have some trouble configuring NAT.
Please check the following link for some usefull information:
http://www.cisco.com/en/US/products/ps6350/products_configuration_guide_chapter09186a008044eddc.html
But answering your question:
straight link:
Avoiding Server Overload Using TCP Load Balancing
Perform this task to configure server TCP load balancing by way of destination address rotary translation. These commands allow you to map one virtual host to many real hosts. Each new TCP session opened with the virtual host will be translated into a session with a different real host.
SUMMARY STEPS
1. enable
2. configure terminal
3. ip nat pool name start-ip end-ip {netmask netmask | prefix-length prefix-length} type rotary
4. access-list access-list-number permit source [source-wildcard]
5. ip nat inside destination-list access-list-number pool name
6. interface type number
7. ip address ip-address mask
8. ip nat inside
9. exit
10. interface type number
11. ip address ip-address mask
12. ip nat outside
http://www.cisco.com/en/US/products/ps6350/products_configuration_guide_chapter09186a008044eddc.html#wp1048769
example:
ip nat pool real-hosts 192.168.15.2 192.168.15.15 prefix-length 28 type rotary
ip nat inside destination list 2 pool real-hosts
interface serial 0
ip address 192.168.15.129 255.255.255.240
ip nat outside
interface ethernet 0
ip address 192.168.15.17 255.255.255.240
ip nat inside
access-list 2 permit 192.168.15.1
All from cisco website.
If this helps, please rate my post.
Thanks,
Vlad

Virtual mail hosts: 255 character limit on SPF records

This one was a surprise to me, and caused a lot of headache, so I thought I'd pass it along.
I'm running multiple virtual mail hosts off of my doughty PowerMac single G5 1.8GHz running OS X Server 10.4.11. Some of the outgoing mail was being bounced as spam because a) there wasn't an SPF record on any of the domains and b) the domain of the mailserver didn't always match the domain of the sender. (Most often, it went out under the hostname of the server, cerberus.limbo.jcf.org—which is useless, since that's a LAN address.)
Trying to be a good citizen (and make sure that all of everyone's mail got through), I added SPF records that explicitly named each and every mailserver on the machine, just so that everything was clear and aboveboard—but they ended up being about 500 characters long.
Fastforward a week or two... and I was having problems with my DNS zones loading—I'd get errors that they'd timed out. After pulling my hair out for a while, I discovered that TXT records have a limit of 255 characters (including spaces, etc.) Some folks running servers on non-OS X Server machines have split the records over multiple TXT records (does that even work?), but you get exactly one TXT record per OS X Server machine: the Comment box.
I've now simplified the SPF records so that they read something like this:
+v=spf1 a mx mx:cerberus.limbo.jcf.org mx:cerberus.jcf.org mx:jcf.org ip:173.164.140.96/30 ip:207.58.140.213/30 include:comcast.businessclass.net include:comcast.com -all+
To translate:
• +v=spf1 a mx+ It authorizes deliveries from any IP listed in the DNS zone, and from any mailserver defined in the zone
• +mx:cerberus.limbo.jcf.org mx:cerberus.jcf.org mx:jcf.org+ It also explicitly authorizes deliveries from the server's main LAN and internet DNS names as well as the domain of the foundation for which I work (and through which emails are occasionally relayed)
• +ip:173.164.140.96/30 ip:207.58.140.213/30+ Next it authorizes the public static IP blocks for the server and the foundation's remote server
• +include:comcast.businessclass.net include:comcast.com+ Finally it includes the domain names of the ISP through which most of the mail are relayed
• -all The last item says that if the mail didn't originate from one of those addresses, it isn't ours.
(I think that I've got that right. If I've botched it anywhere, let me know, okay?)
That's 169 characters. The DNS zones loaded happily, and the mail seems to be going out without getting bounced. So far so good!
(There's probably a way to get the hostname on each email to match the domain from which it is being addressed, but I haven't gotten there yet.)
Message was edited by: David Kudler

Most often, it went out under the hostname of the server, cerberus.limbo.jcf.org—which is useless, since that's a LAN address.
You can control this via the myhostname setting in Postfix. This defines the name it uses to identify itself to remote mail servers, which sounds like it'll address a lot of your issues.
I added SPF records that explicitly named each and every mailserver on the machine, just so that everything was clear and aboveboard—but they ended up being about 500 characters long.
OK, this doesn't make sense. You don't need to list every virtual hostname for every domain.
All you need to do is add this specific mail server's address in each domain.
There's no requirement that the hostname of the mail server matches the domain name, so it's entirely valid to create an SPF record in domain1.com that lists mailserver.someotherdomain.com as authoritative. Then, as long as postfix's myhostname says it's mailserver.someotherdomain.com and your reverse DNS resolves to that address your problem is solved.
...but you get exactly one TXT record per OS X Server machine: the Comment box.
Unless you edit your zone file directly and add whatever other records you like. However, given the above, I don't think the 255-character limit should be an issue.
Even if you didn't want to mess with your zone files directly there's still a way around that - SPF allows for an 'include' record which basically tells remote servers to include the record from some other domain, so for each domain you could just tell it to include some other domain's record (which, in turn, could include another domain) allowing virtually unlimited record length (or, at least, 255 characters per domain you manage).
SPF Includes are covered here.
• include:comcast.businessclass.net include:comcast.com Finally it includes the domain names of the ISP through which most of the mail are relayed
Bzzzz. You've now allowed any other customer of comcastbusiness.net and comcast.com to send mail on your behalf. You probably don't want to do that. When you consider that 'comcast.com' includes every one of their residential customers you can see that you really don't want to do that.

Hosting multiple web sites.

Dear members:
I am considering OS X Server for a personal project and would like some advice to help me decided what route to take.
My computer will soon be turned into a server hosting my web sites and some other services at my home office. I am thinking about OS X Server as I have done some research and found OS X to require quite a bit of work so that it can do virtual hosting.
How simple is it to configure virtual hosting in OS X Server ? I am considering the latest Leopard version in case it matters. I am not experienced and wish to employ an application that will simplify this process.
Any help will be very appreciated.
Thank you,
Joseph Chamberlain

Joseph:
Although I wouldn't touch any Web-server software other than Mac OS X Server, I should warn you that Apple occasionally favors "pretty" over "practical." I've used PCs and Macs since the Mac Plus was hot, and I've used OS X Server 10.3, 10.4, and now 10.5.3. The last is really excellent, with a gorgeous/powerful Server Admin user-interface, many supporting tools, and thousands of hardball features, but like many Apple "industrial-strength" software products, Mac OS X Server still contains a very few really BAD bugs -- or "oversights" if you prefer.
Before taking the plunge (maybe you already did?), read up on troubleshooting OSX Server's Postfix-based mail service. Even the 10.5.3 upgrade can turn "easy setup" into major frustration; you can do everything by the book and still have weird mail-server glitches. Don't even get me started on documentation for and "centralized" management of Open Directory, Kerberos, etc. I will be migrating to SSL mail security to avoid the user-directory thrash.
When trying to make OSX Server 10.5.3 (a clean install) host *multiple top-level Web domains under a single IP number*, Apple documentation online and on disc didn't offer enough setup detail and virtually NO troubleshooting. Google.com can find what's missing. This archived earlier forum thread saved my bacon....
Topic : "multiple web sites"
http://discussions.apple.com/thread.jspa?threadID=682507
In a nutshell: a multiple-website installation of OSX Server requires careful interaction between the domain-name, IP#, port#, and aliases for each site. When I created my several sites. I relied on the default installation (and Apple instructions) which leaves a wildcard "*" asterisk in each site's Web>Sites>Aliases>WebServerAliases field (in Server Admin). No matter how I diddled with each site's domain-name, IP#, and port# (to create unique combinations), I couldn't force the individual domains to get served independently (despite their shared static IP#) until I DELETED the wildcard WebServerAliases for all the sites. Poof, now it works fine.
Part of the problem is that Apple still publishes documentation on its legendary *drool-proof paper* ("do not expose your LaserWriter to open flame").

Advice on running XE on a small virtual Linux host ("Virtuozzo")

Hello,
I'm really excited by XE... I've been developing applications with Oracle at work for a couple of years now and now I can finally get rid of MySQL for the stuff I do in my spare time. Thank you, Oracle! :-)
I've got a virtual Linux root server which is hosted at a local service provider.
They call it a "VPS" (virtual private server), and it runs under Virtuozzo, http://www.sw-soft.com/products/virtuozzo/, which is somehow similar to VMWare.
My idea is to stop using the MySQL database that came preinstalled with the server and install XE instead. Currently, my server is rather limited in terms of available RAM and disk, so I'm considering an upgrade to a bigger type of VPS.
The options most compatible with my budget are:
- a virtual Linux server with 256 MB RAM and 500 MHz CPU or
- a virtual Linux server with 512 MB RAM and 1000 MHz CPU.
Besides a small Oracle XE database, I'll be running Apache with mod_python and a couple of web apps / scripts I'm writing. I don't expect high performance, I just want to have a small Oracle database to play with and to run some web pages.
Does anyone have experience with a similarly limited setup? Will the smaller server still allow some breathing room for other software, or is 256 MB the bare minimum required for Oracle itself?
Any advice is greatly appreciated!

Hello,
I finally got my server upgrade and now have a virtual machine that's "should" be big enough to run XE (256 MB memory usable, 15 GB disk)
However, installation wasn't as smooth as I had hoped... I just can't get it to work. Any ideas? I have no idea what to do next.
Memory stats (these seem to include other machines on the same host):
# free -m -t
total used free shared buffers cached
Mem: 1006 980 25 0 83 430
-/+ buffers/cache: 466 539
Swap: 2000 558 1442
Total: 3006 1538 1467
Installation:
# rpm -i oracle-xe-10.2.0.1-0.1.i386.rpm
This system does not meet the minimum requirements for swap space. Based on
the amount of physical memory available on the system, Oracle Database 10g
Express Edition requires 1509.0 MB of swap space. This system has 1442 MB
of swap space. Configure more swap space on the system and retry the installation.
error: %pre(oracle-xe-10.2.0.1-0.1) scriptlet failed, exit status 1
error: install: %pre scriptlet failed (2), skipping oracle-xe-10.2.0.1-0.1
I tried adding more swap space by creating an additional swap file, but that's not possible; swapon fails with "Operation not permitted". As far as I know, only the hosting machine can add more swap for me.
Using rpm ... --nopre ... does work, though:
# rpm -i --nopre oracle-xe-10.2.0.1-0.1.i386.rpm
Executing Post-install steps..........
usermod: `dba' is primary group name.
You must run '/etc/init.d/oracle-xe configure' as root user to
configure the database.
# /etc/init.d/oracle-xe configure
/etc/init.d/oracle-xe: line 16: /etc/init.d/functions: No such file or directory
Oracle Database 10g Express Edition Configuration
This will configure on-boot properties of Oracle Database 10g Express
Edition. The following questions will determine whether the database should
be starting upon system boot, the ports it will use, and the passwords that
will be used for database accounts. Press <Enter> to accept the defaults.
Ctrl-C will abort.
Specify the HTTP port that will be used for HTML DB [8080]:
Specify a port that will be used for the database listener [1521]:
Specify a password to be used for database accounts. Note that the same
password will be used for SYS, SYSTEM and FLOWS_020100. Oracle recommends
the use of different passwords for each database account. This can be done
after initial configuration:
Confirm the password:
Do you want Oracle Database 10g Express Edition to be started on boot (y/n) [y]:
Configuring Database...
Starting Oracle Net Listener.
oracle-xe 0:off 1:off 2:off 3:on 4:off 5:on 6:off
Starting Oracle Net Listener.
Starting Oracle Database 10g Express Edition Instance.
ps shows that the listener is running, but nothing else. Manually running /usr/lib/oracle/xe/app/oracle/product/10.2.0/server/config/scripts/XE.sh I see this:
ORA-27102: out of memory
Linux Error: 28: No space left on device
select TO_CHAR(systimestamp,'YYYYMMDD HH:MI:SS') from dual
ERROR at line 1:
ORA-01034: ORACLE not available
[... lots of ORA-01034 after that ...]
The css.log looks like this:
Oracle Database 10g CRS Release 10.2.0.1.0 Production Copyright 1996, 2005 Oracle. All rights reserved.
[ OCRUTL][16384]u_set_ocr_error: Parameter was NULL
2005-11-26 19:15:11.509: [ OCROSD][16384]utgdv:2:ocr loc file cannot be opened
2005-11-26 19:15:11.509: [ OCROSD][16384]utopen:1: Couldnt find ocr,[ocrmirror] location in config file
[ OCRUTL][16384]u_set_gbl_comp_error: Parameter was NULL
2005-11-26 19:15:11.509: [ OCRRAW][16384]proprinit: Could not open raw device
2005-11-26 19:15:11.510: [ default][16384]a_init:7!: Backend init unsuccessful : [33]
[ OCRUTL][16384]u_set_ocr_error: Parameter was NULL
2005-11-26 19:15:11.510: [ CSSCLNT][16384]clsssinit: error(33 ) in OCR initialization
LEM stack 1: error [clsdfopen1] [
LEM stack 2: error [clsdfgenfullname] [sclsdgcwd]
Ideas, anyone? Perhaps it's impossible to run XE on this kind of virtual machine...?
Stefan

'we had difficulty reading this feed. host parameter is null' I am getting no where here, help!?

I've had an RSS feed created for an online video podcast but keep getting the message: 'we had difficulty reading this feed. host parameter is null' i've looked at the forums but none of them make any sense. My IT team say the RSS feed is valid and it should work and that is more like down to an itunes error, any help?

This will be our first podcast so do not currently have a page. Below is the feed.
Thank you Roger.
<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<atom:link href="http://dmsukltd.com/rss/rss.xml" rel="self" type="application/rss+xml" />
<title>The Big Picture</title>
<link>http://www.dmsukltd.com/hawk/dms_big_picture/</link>
<description>A brand new monthly movie show produced by DMS</description>
<language>en-gb</language>
<item>
<title>The Big Picture: Mission: Impossible-Ghost Protocol first podcast</title>
<link>http://dmsukltd.com/dl/paramount/tbp_mi4gp_uk_premiere_v1_qt_hires.mov</link>
<guid>http://dmsukltd.com/dl/paramount/tbp_mi4gp_uk_premiere_v1_qt_hires.mov</guid>
<pubDate>15 Dec 2011 12:00:00 GMT</pubDate>
<description>[CDATA[The cast of Mission Impossible - Ghost Protocol joined The Big Picture on the red carpet for the films UK Premiere! Tom Cruise, Simon Pegg, Paula Patton and Samuli Edelmann were on hand to talk about this latest chapter of the popular series!]]</description>
</item>
</channel>
</rss>

Difficulty with Virtual Named hosts on Web Ser

Similar Messages

Maybe you are looking for