Running out of space in /var

[T1homas]

Can any experienced user of RH give me an idea of how he would partition his hda?

During my installation of RH 7.3, I set up /, /var, /home, /dev/shm, /usr/local, /usr, and swap. And also unfortunately I did not check off the "allow the partiton to grow" box, since I'm using a 20 GB HD.

Unfortunately, since I defined my points without enough space, I'm rapidly running out of space in /usr.

Any ideas on how I can increase the partition space of /usr without re-installing the program?

[flipper]

The /usr/local tree should look nearly identical to the /usr tree. Thus you could consider putting all new stuff in /usr/local versus letting it go into /usr. Or consider moving some of the accessory apps from /usr to /usr/local. Other than that... reformat and start again :(

I have a habbit of only letting my Linux server stay put for 3-4 months before a complete reinstall. Perhaps you shouldn't listen to someone like me ;)

[T1homas]

I went ahead and reinstalled RH 7.3 and let the program decide on the partitions. Which leads me to another question.

I have been trying to download various apps that are needed to run the Showeq program and I have seen a large number of packages are listed on the www.redhat.com/errata listing.

Is the command to move them from /errata onto my hda, cp glibc-common 1.1.1 hda?

I'm sorry for my ignornance and basic lack of programming syntax. I do know some of the commands. It's just the way that they are written which appears to be the problem :)

[fryfrog]

try "man rpm" and "man up2date" and "man rhn_register".

as for partitions, if you think you might have space issues on partitions... just have a 32-64mb /boot partition and assisgn the rest to /. it may not be the best when it comes to security or making sure your logs don't keep the entire system from working... but it sure makes deciding on partition sizes easy ;)

[T1homas]

Thank you Fry, for your kind response.

I went ahead and downloaded the RPM's into my /home/<localuser> from the www.rpmfind.net site.

Should I have downloaded them into / instead?

[fryfrog]

the rpms are like .zip files. once you install them, you can delete them. it does not matter where you download them to, nor does it matter where you install them from. mine usually end up in /root or /home/~ or where ever i feel like downloading them to :)

also, go back and run the "rhn_register" command, then use "up2date" to make sure you get your system rpms up to date :)

its pretty easy, i think "up2date -u" will update all the rpms that need updating.

[T1homas]

Thank you for your reply.

I re-logged on in the root directory, and in the terminal emulator mode tyyped this command

rpm -i gbdm-1.8.0-10.i386.rpm

and got the following error message:

failed; No such file or directory.

I downloaded the needed rpm's to my /home/<loccal user> directory and again, I am executing the rpm -i command in the root directory from the terminal emulator.

Where did I go wrong?

[Mr. Suspicious]

I downloaded the needed rpm's to my /home/<loccal user> directory and again, I am executing the rpm -i command in the root directory from the terminal emulator.

That's where you go wrong. You have to execute it AS root (-su) IN "/home/<loccal user> directory"

And just for the record, it's MUCH better to ask general Red Hat related questions in a general Red Hat related Forum and not in a ShowEQ forum, because your question has been asked probably a zillion times there already and you can use the nifty search feature of such a forum to find all kinds of "neat stuff" that will help you.

[T1homas]

Thank you for your kind reply

su root allowed me to get into the root directory from my local user file. Thanks for the heads up :)

The majority of the RPMs installed without any problem. However, I did have two problems:

1. the installation of libstdc ++3.3.0.1-3 gave me the following error message: read manifest failed: success

What does that mean?

I d/l this RPM from a mirror site in Lyon, France because I'm have difficulty accessing rpmfind.net during the daytime hours (I'm on EDT; -5 GMT)

2. the kernel-headers-2.4.7-10 and zlib-1.1.3-24 gave error messages which appeared to be stating that there were conflicts with later versions already installed on the machine.

Do I have to uninstall the newer RPM versions and install these older ones?

Thanks again for all of your kind help in trying help a neophyte get RH 7.3 going with ShowEQ

[high_jeeves]

I'm going to have to go with Mr. S on this one now.. This thread has little, if nothing to do with ShowEQ. You will get better answers if you head over to the RedHat forums, and that will also help keep these forums on topic.

--Jeeves

[T1homas]

I do understand your concern over what seem to be idiotic questions posted by know-nothing posters, but as a complete neo-phyte, I am making progress.

I have gotten all the RPMs installed and have run the shell to get Showeq compiled.

So, please bear with me :) I'm almost there :)

After I typed the ./ configure command, I got two separate error messages, one telling me that I didn't have gcc3 installed and the other telling me that the libEQ.a was not installed in the usr/lib.

After I corrected these two problems, I retyped the shell in the root/seq directory, and got to the make -f Makefile.dist

I then got an error message stating that "bash: make: command not found"

I don't understand it. Where might have I gone wrong?

[high_jeeves]

You still dont have everything installed, please take a look at the walkthrough. ALL of the answers to the questions you have asked so far are answered ATLEAST once in the FAQ or walkthrough.. and atleast 5 more times in threads on the board. Make sure you have all of the RPMs in the list found in the walkthrough and INSTALL.newbies on your system.

--Jeeves

[Dedpoet]

After I corrected these two problems, I retyped the shell in the root/seq directory, and got to the make -f Makefile.dist

Hint: You're in the wrong directory. I'm assuming you made the seq directory from root, and then did a cvs checkout of "showeq"? This means your source code is all sitting under /root/seq/showeq. Try changing to that directory and run your make.

[Yueh]

Somewhat on topic question... Won't parted allow you to resize partitions in place?

[T1homas]

God Bless You Ded !!

I was in the wrong directory. I did the make -f Makefile.dist and it worked !!!!! Thanks :)

However, when I now run the ./ configure I get "bash ./: is a directory"

Aren't you suppose to run the ./configure && make from the /showeq directory?

And, again, I do apologize to all of the experienced users. I realize that a lot of these questions are redundant. However, to someone with no programming experience the process can be daunting.

Thanks again for your hint, Ded !!! :)

[high_jeeves]

READ THE FREAKING WALKTHROUGH! SECTION 1.9!

It tells you in NO UNCERTAIN terms which directory, exactly what commands to type (note, dont add your own spaces to the commands just for the fun of it, type them as they are in the walkthrough).

This is rediculous!!! If you tell us your address, maybe one of us will come to your house, but until then, try helping yourself out a bit by reading before you ask. You would already have the whole thing set up by now if you tried to answer your own questions before you posted!



And, again, I do apologize to all of the experienced users. I realize that a lot of these questions are redundant. However, to someone with no programming experience the process can be daunting.


How daunting is following instructions? They are there, in black and white.. just read them, and follow them.. you can even cut the command and paste it, so you dont actually have to do any typing...

--Jeeves

[T1homas]

Dear Ded,

I was able to run the /configure from the /root/showeq directory and start the compile.

However, I have now received the following error message during the compile

"configure was unable to locate a MOC binary anywhere on your system!

If you have a working MOC binary, please set the environment variable MOC to point to the location of your MOC binary and configure over"

I'll continue to research this issue on "MOC". But thanks again for getting me this far !! :)

Thanks again :)

[high_jeeves]

ROFL. If you would simply follow the directions in the walkthrough (section 1.9) you would not get this error. If you are being too bullheaded to do this correctly, then you will continue to get that error.

--Jeeves

[T1homas]

Where, specifically, is the Walkthrough 1.9?

Is it one of the README.TXT in the /root/showeq directory?

Look, I'd love to compile this thing and be done with it. But again, for your average person with no programming experience, the process is a bit confusing.

Look at the numbers. There are about 400,000 EQ users. Out of that universe, there are maybe 2,000 downloads of the libEQ.a.

If the darn process was so straightforward, don't you think you'd have a much great number of downloads?

I do appreciate all of the help I have received here on the boards; I sincerely do.

Thanks to all of the kind people and straight forward posts from them.

If you want to go ahead and laugh--go ahead. But it does reflect somewhat on your own persona.

I'm just trying, perhaps not as smoothly as others, to get the thing running.

Thanks again :)

[Mr. Suspicious]

For the Faq: See the link on the main page of this Forum, it's in BIG typefont.

For the "walk-through" read the INSTALL.newbies (http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/seq/showeq/INSTALL.newbies?rev=1.3&content-type=text/vnd.viewcvs-markup) file found in your showeq directory. Read the documentation (don't skimm over it, actually read it), it's there for a reason.

[T1homas]

Dear Mr. S

Thanks for all of your help. I have read it and I'm now wondering which of the required RPMs (if any) I might be missing.

Isn't there a find command that would let me query my hda to find out which packages are already there?

Is it -qa [filename] ?

Thanks again for your kind reply

[Mr. Suspicious]

I seriously wouldn't know how to do that, I never had to. Just for giggles. I went to http://www.redhat.com (rpm = RedHat Package Manager) and typed in "find out which packages are already there" (exactly your words) in the searchbox and pressed enter.

The 2nd result in the list: http://www.redhat.com/docs/manuals/linux/RHL-5.0-Manual/user-guide/doc061.html (titled: Using RPM) tells us:



8.2.4 Querying
Querying the database of installed packages is accomplished with rpm -q. A simple use is rpm -q foo which will print the package name, version, and release number of the installed package foo:


$ rpm -q foo
rpm-2.0-1

Instead of specifying the package name, you can use the following options with -q to specify what package(s) you want to query. These are called Package Specification Options.


-a queries all currently installed packages.
-f <file> will query the package owning <file>.
-F is the same as -f except it takes filenames via stdin
(e.g. find /usr/bin | rpm -qF).
-p <packagefile> queries the package <packagefile>.
-P is like -p except it takes package filenames from stdin
(e.g. find /mnt/cdrom/RedHat/RPMS | rpm -qP).
There are a number of ways to specify what information to display about queried packages. The following options are used to select the information you are interested in. These are called Information Selection Options.


-i displays package information such as name, description, release, size, build date, install date, vendor, and other miscellaneous information.
-l displays the list of files that the package ``owns''.
-s displays the state of all the files in the package. There are only two possible states, normal and missing.
-d displays a list of files marked as documentation (man pages, info pages, README's, etc).
-c displays a list of files marked as configuration files. These are the files you change after installation to adapt the package to your system (sendmail.cf, passwd, inittab, etc).
For those options that display file lists, you can add -v to your command line to get the lists in a familiar ls -l format.


So it's



[]#rpm -q -a

or

[]#rpm -qa


to get a full listing of all packages installed, or




[]#rmp -q <packagename>


to check if a specific package is installed.

You know, there are places full of wealthy information (the manual of your distro is ammongst the first places to look for ANY information about Linux), take the time to visit them. Also don't be afraid to experiment. It's not like you can FUBAR without ever being able to reinstall.

[high_jeeves]

Look at the numbers. There are about 400,000 EQ users. Out of that universe, there are maybe 2,000 downloads of the libEQ.a.

If the darn process was so straightforward, don't you think you'd have a much great number of downloads?



Holy misinformation Batman!

1) You need two machines, so that cuts the 400,000 base down to about oh, 4000 people. You need to run linux, our number drops to 2000.

2) WE DONT WANT THE PROCESS TO BE EASY! WE WANT IT TO BE HARD! If you aren't bright enough to read, and follow basic directions, then you arent the type of person that should be given the additional power of ShowEQ. The devs have gone out of there way to make things more difficult, to this exact means.

Why do you keep talking about programming experience? You have to have exactly 0 programming experience to install ShowEQ, you do exactly 0 programming during this process. Many people with 0 linux experience have read the FAQ and Walkthrough (NOTE: they are pointed out in large type on the front page) and been able to install ShowEQ without posting a single question. The problem here is, you simply refuse the do some work for yourself, but rather expect that we should come running to help you. Please, read the walkthrough, follow it step by step, it is as simple or complex of a process as you want to make it. For the complete neophyte it should take about 1 hour + compile time. If it is taking longer, that is because you simply wont stop asking questions long enough to read the answers that are already in the walkthrough and FAQ. Please spend 20 minutes reading the available information, and then try to install.

--Jeeves

[Dedpoet]

T1homas, did you export your QTDIR and PATH variables before you did the make?

The walkthru is here (http://seq.sourceforge.net/showthread.php?s=&threadid=273&perpage=15&pagenumber=1), but it's for Rehat 7.2.

Once you get the configure done, you should be doing something like this:

export QTDIR=/opt/qt-gcc3-2.3.2 (this may be different for you depending on where QT is installed for you)
export PATH=$QTDIR/bin:$PATH
export MANPATH=$QTDIR/doc/man:$MANPATH
export LD_LIBRARY_PATH=$QTDIR/lib:$LD_LIBRARY_PATH
export CXX=g++3
export CC=gcc3
make -f Makefile.dist
./configure
make
make install

If you are still stuck, you can PM me. I have nothing but respect for Jeeves and Mr. Suspicious, but if you keep posting here instead of reading the FAQ and walkthru that I linked, they will continue to flame you. You are very close to a working system, and I would really like to see this thread end, so let's get this solved.