I upgraded some days ago from F14 to F15, I once had postgresql-server configured and running, but now I'm getting errors when starting the service and I can't understand where the issue is.
Here some infos
Code: # /etc/init.d/postgresql start Starting postgresql (via systemctl): Job failed. See system logs and 'systemctl status' for details. [FAILED] Code: # systemctl --failed UNIT LOAD ACTIVE SUB JOB DESCRIPTION
I have two machines. One is for web server and the other is for database server.I need to install php in the web server and postgresql in the database server. I installed postgresql. But could not install php as it needs postgresql library during configuration steps.
I'm wondering how I start it automatically when I boot up the ubuntu 10.04? especially, when I created the user "postgres", I set a password to it. every time I issued "su - postgres" command, I need to give it a password.
I'm trying to install pg_config so I can get some simple reports regarding my PostgreSQL 8.4.7 installation.
I downloaded pg_config and its dependencies and tried to install them with the package installer, but it wouldn't go because it seems like something may be wrong with the libkrb5-3 and libkrb5support0 packages - they won't install and the package installer just zips through without reporting missing dependencies or anything. The MD5, SHA1, and SHA256 checksums all match, but they just won't install.
I downloaded all the packages from the ubuntu [security] repository.
Because my Ubuntu 10.04 (Lucid) machine is not connected to the internet, I download packages to my Windows 7 machine, transfer them across my network to a shared folder on my Ubuntu machine and double-click the package to install.
I've successfully installed postgresql-8.4, apache-2.2.14, php-5.3.2, drupal-6.16, and several other packages and all their dependencies this way without any problems before.
To make sure I wasn't having some new problem, I then tried installing libk5crypto3 - it installed fine.
I sent an email to frank@lichtenheld.de , the repository maintainer, but haven't heard back from him yet.
I upgraded my server machine to F12 without realising that postgresql gets upgraded to 8.4, and without realising that 8.4 can't work with 8.3 databases.
Is there a way to install postgresql 8.3 on F12, preferably from Fedora repositories?
Alternatively, is there a way to migrate an 8.3 database to 8.4? I know that the recommended approach is to dump and restore the DB, but now that I have upgraded to F12, the pg_dump command doesn't work anymore.
There is also this open-source tool called pg_migrator but I couldn't find a package for it and it requires a complicated compilation and installation procedure.
The simplest solution seems to me to remove pg 8.4 and install 8.3 on Fedora 12.
Is there any simpler method to downgrading to PostgreSQL 8.2.x? I am installing an application that requires it. I have removed postgresql 8.3 and tried to install some older fedora packages, but it did not go well.
My next step was to compile from scratch. However, I'm not real experienced with make/compiling and am fond of the organizational structure of RPMs.
The other option I could pursue is keeping 8.3 and installing an isolated 8.2 server for exclusive use by this app.
I'm still new to PHP and SQL, but all the tutorials I've found connect to the database like this:
PHP Code:
Right now I'm just doing local network tests before exposing everything to the Internet.
Wouldn't leaving the password in there as plain text be a huge security issue? I tried downloading the php file off the server, and it just comes down blank. So does PHP already have a security feature that doesn't allow anyone to just nab PHP files off the server?
And for Postgre, I have pg_hba.conf set up to "trust" it's own IP address:
Code:
Would I need to use something like Kerberos, PAM, or ident authentication? Right now the only plan is to use it as login system for a website. The clients themselves won't be accessing the DB itself, because all the DB access will be through PHP.
When I recently upgraded my desktop to Karmic, my postgresql 8.3 no longer starts. I've got data in 8.3 that I need, but now no way to access it. Karmic upgrades to Postgresql 8.4, but leaves 8.3. Problem is, I can't seem to start 8.3 no matter what I do (there's no error message. I even tried removing 8.4 entirely--without effect. When I try to start 8.3 (using command line or boot-up manager), there is no error message, either in system logs or in the postgresql log file. There is also no indication of an active postmaster in ps aux or anything listening to postgresql ports in nmap (for that matter, seems there's nothing listening even with 8.4 activated).
I could try to 'upgrade' my data cluster to 8.4. Never tried that before and am concerned about possibly losing data. Also, I made a load of tuning adjustments to 8.3 that I'd have to replicate with 8.4. I'd rather stay with 8.3, if possible. There are also two pieces of software, postgresql-common and postgresql-client-common, that presumably allow one to run two versions of postgresql simultaneously. Maybe this stuff is even the source of my start up problems. No idea where to learn how to use this software, and, esp., whether this would solve the 8.3 start problem.
Web Server ================ I have been successful to set up the pic-1 server. | | This requires only one machine. | Apache |
[code]....
But now I have two machines.One for Database server and the other is for Web server.pic-2 describes it.
Server Database Server ================ ================ | | | | | Apache | | |
[code]....
I have installed postgresql in one machine.Apache has also been installed in the other machine. But php requires postgresql library support during configuration.I cant remotely access the postgresql from the other machine during configuration.
I have recently Installed Fedora 11. I have installed postgres like yum install postgresql but I wonder there is no user created for postgres when I tried su postgres system says, no identity for user postgres. even when I tried to connect postgres server using psql then it says
Code:
psql: could not connect to server: No such file or directory
Is the server running locally and accepting connections on Unix domain socket "/tmp/.s.PGSQL.5432"?
I just upgraded a Ubuntu Desktop 64bit system from Karmic (9.10) to Lucid (10.04). There were several PostgreSQL 8.3 clusters running on that machine. After the upgrade, postgresql was not able to run anymore.
The package postgresql-8.3 was marked "rc" in dpkg -l I was not able to re-install PostgreSQL 8.3 because of a mismatched dependency. There was no warning. I guess i should have read the release notes, which i didn't. There should have been a warning before it was too late to turn back and upgrade postgresql to 8.4 first. I don't see how this is supposed to work, ever.
I'd like to know if this would be considered a bug, if it is known already and how to correctly file this bug correctly if appropriate (ubuntu-bug ?)
this is my first post right here. I'm running on my first Fedora full experience, then I'm not so experienced about the specific details of the distro. Coming from Ubuntu (and certainly sad about Unity bugs), I wanted to know other distros, so I've adopted Fedora as my main OS. I often had some bugs and try to resolve them by myself but, this time, I've no idea about what can I do. So, there it goes:
Recently I've been trying to set up my PostgreSQL server for development purposes. Then I went to EnterpriseDB website and downloaded the x86-64 GUI install bundle, as I always did in Ubuntu (I never had this bug with Ubuntu). After the download, I went on with the installation, doing these steps:
I run the installation with no problem, doing it alright, then I finish the installation and open pgAdmin3, for set up my databases. I log onto my server, put my password for postgres user and create a new server. It's all OK, so when I go to Query Mode tool (Ctrl + E), it crashes along with the pgAdmin.I decided to run the pgadmin script on terminal, looking for more details, then:
Code:
/opt/PostgreSQL/9.0/pgAdmin3/bin/pgadmin3: /opt/PostgreSQL/9.0/lib/libuuid.so.1: no version information available (required by /usr/lib64/libSM.so.6) /opt/PostgreSQL/9.0/scripts/launchpgadmin.sh: line 3: 4890 Segmentation fault (core dumped) LD_LIBRARY_PATH=/opt/PostgreSQL/9.0/pgAdmin3/lib:/opt/PostgreSQL/9.0/lib:$LD_LIBRARY_PATH G_SLICE=always-malloc /opt/PostgreSQL/9.0/pgAdmin3/bin/pgadmin3
In our development box we are configured postgresql to work with Jboss. The thing is we have firewall iptables in our linux box. when the iptables is stopped we can connect the postgres db locally using -h option, also we can connect thru a weblink we have created using Jboss. But when the iptables is started we can't connect the db locally using the -h option and the web is giving the below error. Caused by:
org.postgresql.util.PSQLException: Connection refused. Check that the hostname and port are correct and that the postmaster is accepting TCP/IP connections. at org.postgresql.core.v3.ConnectionFactoryImpl.openC
I have weird problem with postgresql + nginx + php5 0 ubuntu 11.04. Postgresql works just fine, when I run "php test.php" from shell. I assume php runs from shell are called cgi runs? When I open page from browser [URL]... I get following error. I assume php runs from browser are called cgi runs? Fatal error: Call to undefined function pg_connect() in /var/www/mydomain/htdocs/test.php on line 7 When you print php info, only reference to postgresql is pgsql.ini and pdo_pgsql.ini files that are parsed.
[code]....
I did install postgresql by running following commands:
Downloaded few minutes ago Fedora 15 live x64 and used Fedora live usb creator to make my usb drive bootable.Once configured correctly my bios Syslinux starts but it hangs on the startup message
I'm attempting to install F11 on my desktop. I am unable to complete the installation because the X environment for anaconda fails to start and instead the installation reverts to text mode (in which I cannot seem to to a custom disk partitioning layout for the installation, which I have to be able to do). This was not a problem with the Alpha or Beta, but the preview and now the final release have both had this problem.
My system: intel 3.16 core 2 duo 4gb ddr2 ram dual nvidia 9800gtx+
I have just installed F10 via DVD download onto my Dell Desktop with Nvidia GE6800 graphics card and twin LCD Monitors. I have configured Dual Boot with Grub (Windows XP Home and F10). After installing and configuring the correct driver from Nvidia I am now observing a strange problem when booting into Fedora.
On cold boot the system starts and gets to Probing EDD (edd=off to disable)... and then freezes, however if I then warm boot Ctrl Alt Del the system boots into F10 with no problem. I had to disable the latest Kernel update because I couldn't start graphical interface with it so I only boot into the original Kernel.
Posting from Splashtop as i can't get past keytouch-apid in the boot process,Yeah me bad so i would like to disable it so i can complete the process.Using the interactive startup seemed the way to go but how does one get there before udev starts up?Or is there something else i could try.
Some quick background info: I am not a Linux expert, but I'm not sure I can still qualify as a "newbie", since my first Linux install was RH9 in 2003. I later upgraded to FC3 and continued to use it until my hard drive crashed around a year ago. I suffered on an old WinXP machine until I was able to buy a new system from someone on Craig's list. Although the guy didn't realize it (hence neither did I), it has a 64-bit processor, so when I installed Fedora 10, I installed the x86_64 version.I chose F10 since I use the Planet CCRMA pro audio packages, and they lag slightly behind the latest and greatest distro versions. I'm also not a bleeding-edge kinda guy, since - as I said above - I am not an expert.
So, then, here's what's happened: I installed F10 by downloading the 64-bit DVD image from the Fedora Unity respin site. As I had hoped, after the installation when I asked yum if I was due for an upgrade it said I was up to date.That was fine for a few weeks, but then this past Sunday I got a pop-up message saying that I was due for a system upgrade. I was hesitant, as I usually am before making any drastic changes, but I figured this is Fedora, they want to do the upgrade, this has got to be pretty safe. Not so much.I fired off the upgrade before heading to bed and let it run overnight (it actually took only an hour or two, but I started it late). In the morning I found my system with a blank screen and essentially unresponsive. I can move a block cursor around the screen, and if I type anything it is echoed on the screen, but there is no response; I don't believe I am in a shell of any kind.
Hitting CTRL-ALT-DEL reboots the system, and it gets past the grub menu and the Fedora start-up progress bar (the dark blue/light blue/white text blocks with "Fedora 10" at the right). After the progress bar finishes, though, the screen goes blank (maybe X is starting up?) but never gets anywhere. That's when I can type stuff on the screen, but nothing provokes any kind of response (except for CTRL-ALT-DEL).I'm mildly curious to know if the recommendation about upgrades vs. fresh installs, as given in the sticky post on this forum, still applies, or if it is now considered safer to let the system upgrade itself when it feels ready.
What I really need to know, however, is where I should start in order to get my system back. Should I go to the command line from the grub menu? Boot from my install disc in rescue mode? I'm not even sure what I should look for once I've done either of these, so I'm kind of floundering ...This PC has a dual-core Pentium 4 running at 2.9 GHz, and apparently it's a 64-bit processor. The video card is by NVidia, but I don't know the exact card off-hand. There's a built-in Intel sound card of some kind and I added an M Audio Delta 1010LT multi-channel card.
I download the Fedora 12 and updated from Fedora 11. After installing successful and restart, I can't find Fedora 12 selection on the Grub. I try to write it on the grub.conf, but it's still can't start the Fedora 12. code...
I can connect to localhost, and get the test page.However, as soon as I install the apache PHP module, with:[root] # yum install php apache will not restart.Watching active processes with qps, I see apache tries to start, but then it crashes - hence it writes to the error log to say it's started. But there's no PID in /var/run/httpd, and I cannot connect to localhost.
I have an Intel Core 2 Duo board with Intel GS45 + Intel ICH9M chipset. After installed from DVD (i386), it will ask reboot, and I clicked reboot button, the system reboots, but it will freeze somewhere from 3-threaded progress bar to the desktop show up. It freeze most likely in the "F"-icon flash or the dots showing up.
I downloaded fedora from the site, burned the iso it starts up, the loading fedora logo appears and after that I get a strange grey and black screen and no login screen.
I have win 7 running on this laptop 2gb ram, 1,86ghz... 256 graphic card.
I have been using Fedora 12/13 for last 6 months or so but still I never tried my hands with the customization of its desktop and its kernel which is the beauty of Linux so I want some links or guidelines about how to start doing something meaningful with Fedora.
I just installed fedora on an old laptop. After restarting, the computer just shows me a black screen, it doesn't even show the BIOS screen. If I put a CD with the fedora installation program, it sounds like its reading it, but the screen keeps being black.
Just as background, i thought this laptop had died 4 months ago when I tried to start it and then nothing happened (black screen). Today, I decided to check if it was working before throwing it to the garbage, with the fedora 13 live cd in it, and voila, in entered and I was able to install fedora. But then upon installation of fedora it just shows again the same black screen...can this be a hardware problem?
