disk full? how should handle this?

disk full? how should handle this?

Post by Gene » Mon, 18 Mar 2002 15:12:03



I was trying to install j2sdk1.4.0 to /usr/j2se, but in the middle of the
installation, it says disk full.

What does it mean?  Does it mean the partition of /usr is full?  or is
/usr/j2se full?  How can I handle this problem?  Do I have to remove
something from /usr?  I dare not.

 
 
 

disk full? how should handle this?

Post by Dennis Clark » Mon, 18 Mar 2002 15:45:49



> I was trying to install j2sdk1.4.0 to /usr/j2se, but in the middle of the
> installation, it says disk full.

> What does it mean?  Does it mean the partition of /usr is full?  or is
> /usr/j2se full?  How can I handle this problem?  Do I have to remove
> something from /usr?  I dare not.

It sounds like you have filled up the /usr filesystem.  Worse than that is the
possibility that you did while acting as the root user.  That probably means
that you have used up all available filesystem space as well as the small
percentage that is typically reserved on the filesystem.  Hopefully you can
backout the install.  Is there a README file that came with the package?  If
not then you have the * task of removing the /usr/j2se directory tree and
do so carefully.  Ensure that you do not remove anything from the other areas
of /usr.  Its possible to run a pkgchk on every package installed in the
system but this would be a very time consuming and boring task.  It would tell
you about every problem.  

In the future, after you have cleaned this problem up, use df -ak to check for
available disk space and always have a clean backup before diving in with
fingers crossed for good luck.  I have a post about this sort of thing coming
up shortly.   In any case, I'd suggest that you create a new filesystem and
mount it at /usr/j2se such that you won't have this problem in the future.

Dennis

 
 
 

disk full? how should handle this?

Post by Gene » Mon, 18 Mar 2002 16:31:43


Sounds complicated to me.  So do you mean that I would not have used up the
space had I not operated as the root user?

And removing a directory like j2se is not as tedious as you said, just do
rm -r j2se, right?

I did pkgrm SUNWj2* before I tried installing jdk1.4.0.  There is no way I
can expand the /usr partition?

And you  suggest that I create a new filesystem and mount it at /usr/j2se,
what do you mean?  You want me to do rm -r /usr and create /usr via mkdir
usr and then mkdir j2se?  I dare not.  There are tons of files under /usr.

Sounds like I ran into big trouble?




> > I was trying to install j2sdk1.4.0 to /usr/j2se, but in the middle of
the
> > installation, it says disk full.

> > What does it mean?  Does it mean the partition of /usr is full?  or is
> > /usr/j2se full?  How can I handle this problem?  Do I have to remove
> > something from /usr?  I dare not.

> It sounds like you have filled up the /usr filesystem.  Worse than that is
the
> possibility that you did while acting as the root user.  That probably
means
> that you have used up all available filesystem space as well as the small
> percentage that is typically reserved on the filesystem.  Hopefully you
can
> backout the install.  Is there a README file that came with the package?
If
> not then you have the * task of removing the /usr/j2se directory tree
and
> do so carefully.  Ensure that you do not remove anything from the other
areas
> of /usr.  Its possible to run a pkgchk on every package installed in the
> system but this would be a very time consuming and boring task.  It would
tell
> you about every problem.

> In the future, after you have cleaned this problem up, use df -ak to check
for
> available disk space and always have a clean backup before diving in with
> fingers crossed for good luck.  I have a post about this sort of thing
coming
> up shortly.   In any case, I'd suggest that you create a new filesystem
and
> mount it at /usr/j2se such that you won't have this problem in the future.

> Dennis

 
 
 

disk full? how should handle this?

Post by Gene » Mon, 18 Mar 2002 16:36:55


My current disk usage condition is like this (copy-paste to notepad and use
courier new to achieve a better view of the table):

# df -ak
Filesystem            kbytes    used   avail capacity  Mounted on
/dev/dsk/c0t0d0s0    1488798 1445823       0   100%    /
/proc                      0       0       0     0%    /proc
fd                         0       0       0     0%    /dev/fd
mnttab                     0       0       0     0%    /etc/mnttab
swap                  794712       8  794704     1%    /var/run
swap                  795032     328  794704     1%    /tmp
/dev/dsk/c0t0d0s7    12676347 2222097 10327487    18%    /export/home
-hosts                     0       0       0     0%    /net
auto_home                  0       0       0     0%    /home
-xfn                       0       0       0     0%    /xfn
unknown:vold(pid242)       0       0       0     0%    /vol

But I don't really understand those data.




> > I was trying to install j2sdk1.4.0 to /usr/j2se, but in the middle of
the
> > installation, it says disk full.

> > What does it mean?  Does it mean the partition of /usr is full?  or is
> > /usr/j2se full?  How can I handle this problem?  Do I have to remove
> > something from /usr?  I dare not.

> It sounds like you have filled up the /usr filesystem.  Worse than that is
the
> possibility that you did while acting as the root user.  That probably
means
> that you have used up all available filesystem space as well as the small
> percentage that is typically reserved on the filesystem.  Hopefully you
can
> backout the install.  Is there a README file that came with the package?
If
> not then you have the * task of removing the /usr/j2se directory tree
and
> do so carefully.  Ensure that you do not remove anything from the other
areas
> of /usr.  Its possible to run a pkgchk on every package installed in the
> system but this would be a very time consuming and boring task.  It would
tell
> you about every problem.

> In the future, after you have cleaned this problem up, use df -ak to check
for
> available disk space and always have a clean backup before diving in with
> fingers crossed for good luck.  I have a post about this sort of thing
coming
> up shortly.   In any case, I'd suggest that you create a new filesystem
and
> mount it at /usr/j2se such that you won't have this problem in the future.

> Dennis

 
 
 

disk full? how should handle this?

Post by Dennis Clark » Mon, 18 Mar 2002 17:00:11



> My current disk usage condition is like this (copy-paste to notepad and use
> courier new to achieve a better view of the table):

> # df -ak
> Filesystem            kbytes    used   avail capacity  Mounted on
> /dev/dsk/c0t0d0s0    1488798 1445823       0   100%    /
> /proc                      0       0       0     0%    /proc
> fd                         0       0       0     0%    /dev/fd
> mnttab                     0       0       0     0%    /etc/mnttab
> swap                  794712       8  794704     1%    /var/run
> swap                  795032     328  794704     1%    /tmp
> /dev/dsk/c0t0d0s7    12676347 2222097 10327487    18%    /export/home
> -hosts                     0       0       0     0%    /net
> auto_home                  0       0       0     0%    /home
> -xfn                       0       0       0     0%    /xfn
> unknown:vold(pid242)       0       0       0     0%    /vol

> But I don't really understand those data.

I think that your situation is far more simple than the mess that I have here
with GNOME 1.4.  The aove table essentially states that you have two disk
partitions.  They are on c0t0d0s0 and c0t0d0s7.  It sounds like you don't have
a lot of admin experience so I'm surprised that you tackled the install.  Then
again, I've done far worse on other peoples equipment.  Years and years ago.
I have learned a lot since the 80's.  In any case, it is the middle of the
night and I have nothing better to do than listen to old Nazareth and remove
GNOME from my Ultra 2 so let me see if I can be of help here.

You have a SCSI controller whose number is zero, called c0.  On that
controller you have a disk whose SCSI target or ID number is also zero.  It
also has a logical unit number of zero.  That means that you have a disk
called c0t0d0.  That disk, like all disks, is round.  Yep.  It may be in a
square box but it is round.  I'm messing with ya so just chill.  That disk has
a pile of cylinders numbered from 0 through to a high number.  I hope.  Do a
prtvtoc /dev/rdsk/c0t0d0s0 and letme know what you have there.  Looks like the
/export/home filesystem is just swimming in space.  We need to reorganize your
disk.

Do you have a tape drive?  Please say yes.

We need to move some things around in order to get you up and running.

Dennis

 
 
 

disk full? how should handle this?

Post by Gene » Tue, 19 Mar 2002 02:00:16


Thanks a lot.  Here is what I got from prtvtoc:

# prtvtoc /dev/rdsk/c0t0d0s0
* /dev/rdsk/c0t0d0s0 partition map
*
* Dimensions:
*     512 bytes/sector
*      63 sectors/track
*      16 tracks/cylinder
*    1008 sectors/cylinder
*   29651 cylinders
*   29649 accessible cylinders
*
* Flags:
*   1: unmountable
*  10: read-only
*
*                          First     Sector    Last
* Partition  Tag  Flags    Sector     Count    Sector  Mount Directory
       0      2    00    1049328   3074400   4123727   /
       1      3    01          0   1049328   1049327
       2      5    00          0  29886192  29886191
       7      8    00    4123728  25762464  29886191   /export/home
#

And unfortunately, I don't have a tape drive.  But can I just install the
JDK to whereever there is enough disk space and make a symbolic link to
/usr/j2se?




> > My current disk usage condition is like this (copy-paste to notepad and
use
> > courier new to achieve a better view of the table):

> > # df -ak
> > Filesystem            kbytes    used   avail capacity  Mounted on
> > /dev/dsk/c0t0d0s0    1488798 1445823       0   100%    /
> > /proc                      0       0       0     0%    /proc
> > fd                         0       0       0     0%    /dev/fd
> > mnttab                     0       0       0     0%    /etc/mnttab
> > swap                  794712       8  794704     1%    /var/run
> > swap                  795032     328  794704     1%    /tmp
> > /dev/dsk/c0t0d0s7    12676347 2222097 10327487    18%    /export/home
> > -hosts                     0       0       0     0%    /net
> > auto_home                  0       0       0     0%    /home
> > -xfn                       0       0       0     0%    /xfn
> > unknown:vold(pid242)       0       0       0     0%    /vol

> > But I don't really understand those data.

> I think that your situation is far more simple than the mess that I have
here
> with GNOME 1.4.  The aove table essentially states that you have two disk
> partitions.  They are on c0t0d0s0 and c0t0d0s7.  It sounds like you don't
have
> a lot of admin experience so I'm surprised that you tackled the install.
Then
> again, I've done far worse on other peoples equipment.  Years and years
ago.
> I have learned a lot since the 80's.  In any case, it is the middle of the
> night and I have nothing better to do than listen to old Nazareth and
remove
> GNOME from my Ultra 2 so let me see if I can be of help here.

> You have a SCSI controller whose number is zero, called c0.  On that
> controller you have a disk whose SCSI target or ID number is also zero.
It
> also has a logical unit number of zero.  That means that you have a disk
> called c0t0d0.  That disk, like all disks, is round.  Yep.  It may be in a
> square box but it is round.  I'm messing with ya so just chill.  That disk
has
> a pile of cylinders numbered from 0 through to a high number.  I hope.  Do
a
> prtvtoc /dev/rdsk/c0t0d0s0 and letme know what you have there.  Looks like
the
> /export/home filesystem is just swimming in space.  We need to reorganize
your
> disk.

> Do you have a tape drive?  Please say yes.

> We need to move some things around in order to get you up and running.

> Dennis

 
 
 

disk full? how should handle this?

Post by Solaris X. Migh » Tue, 19 Mar 2002 08:43:20


Is there a need for such a big /export/home FS?  It appears that you
are only using about 14GB of the disk.  Since I don't know that type of box
this is, I will guess that the disk is a 18GB (or bigger?).  If so, you should
be able to use the last couple of gigs and create, say, a /usr FS, boot up
the box to single-user from CD-ROM, move some stuff around and you should be
all set.

Before we go in-depth though, I would suggest looking at /export/home and if
you have any free space.  If the size if not needed, save the stuff or move
it to a new partition and you will be in a much better sharp later when
you need more space.

I would also suggest grabbing the Solaris docs from docs.sun.com to take a
look.  You seem to need it...

Orphy! :-)



> > My current disk usage condition is like this (copy-paste to notepad and use
> > courier new to achieve a better view of the table):

> > # df -ak
> > Filesystem            kbytes    used   avail capacity  Mounted on
> > /dev/dsk/c0t0d0s0    1488798 1445823       0   100%    /
> > /proc                      0       0       0     0%    /proc
> > fd                         0       0       0     0%    /dev/fd
> > mnttab                     0       0       0     0%    /etc/mnttab
> > swap                  794712       8  794704     1%    /var/run
> > swap                  795032     328  794704     1%    /tmp
> > /dev/dsk/c0t0d0s7    12676347 2222097 10327487    18%    /export/home
> > -hosts                     0       0       0     0%    /net
> > auto_home                  0       0       0     0%    /home
> > -xfn                       0       0       0     0%    /xfn
> > unknown:vold(pid242)       0       0       0     0%    /vol

> > But I don't really understand those data.

> I think that your situation is far more simple than the mess that I have here
> with GNOME 1.4.  The aove table essentially states that you have two disk
> partitions.  They are on c0t0d0s0 and c0t0d0s7.  It sounds like you don't have
> a lot of admin experience so I'm surprised that you tackled the install.  Then
> again, I've done far worse on other peoples equipment.  Years and years ago.
> I have learned a lot since the 80's.  In any case, it is the middle of the
> night and I have nothing better to do than listen to old Nazareth and remove
> GNOME from my Ultra 2 so let me see if I can be of help here.

> You have a SCSI controller whose number is zero, called c0.  On that
> controller you have a disk whose SCSI target or ID number is also zero.  It
> also has a logical unit number of zero.  That means that you have a disk
> called c0t0d0.  That disk, like all disks, is round.  Yep.  It may be in a
> square box but it is round.  I'm messing with ya so just chill.  That disk has
> a pile of cylinders numbered from 0 through to a high number.  I hope.  Do a
> prtvtoc /dev/rdsk/c0t0d0s0 and letme know what you have there.  Looks like the
> /export/home filesystem is just swimming in space.  We need to reorganize your
> disk.

> Do you have a tape drive?  Please say yes.

> We need to move some things around in order to get you up and running.

> Dennis

 
 
 

disk full? how should handle this?

Post by zg.. » Tue, 19 Mar 2002 12:27:57



Quote:> Is there a need for such a big /export/home FS?  It appears that you
> are only using about 14GB of the disk.  Since I don't know that type of box
> this is, I will guess that the disk is a 18GB (or bigger?).  If so, you should
> be able to use the last couple of gigs and create, say, a /usr FS, boot up
> the box to single-user from CD-ROM, move some stuff around and you should be
> all set.
> Before we go in-depth though, I would suggest looking at /export/home and if
> you have any free space.  If the size if not needed, save the stuff or move
> it to a new partition and you will be in a much better sharp later when
> you need more space.
> I would also suggest grabbing the Solaris docs from docs.sun.com to take a
> look.  You seem to need it...
> Orphy! :-)

Actually, everything was installed by default.  I did not intervene.  So is it possible to
squeeze the /export/home/ FS and give a little bit to something else?

I don't even know the size of my hard disk.  I run df, du, or things like that, but got
confused by their outputs. :(

The sun site doc is a good resource, but it's too huge, hard to dig out what I want. If
you guys can recommend a good book, I would appreciate it.

 
 
 

disk full? how should handle this?

Post by Federico Balb » Fri, 29 Mar 2002 00:42:31



> And removing a directory like j2se is not as tedious as you said, just do
> rm -r j2se, right?

yup

Quote:> And you  suggest that I create a new filesystem and mount it at /usr/j2se,
> what do you mean?  You want me to do rm -r /usr and create /usr via mkdir
> usr and then mkdir j2se?  I dare not.  There are tons of files under /usr.

it means that you can for example add a new disk and mount it under
/usr/j2se in the /etc/vfstab file. To me the easiest way for you is to
create a link to a free partition..

Do a df -k and look where you have some space available (-k will show in
kilobytes). let's say you have plenty of space in /opt for example.
You may create your j2se under /opt with mkdir /opt/j2se
Then go in /usr and create a link to /opt/j2se like this:

cd /usr
ln -s /opt/j2se j2se

this will create a link under /usr so everytime you write on /usr/j2se the
files will be physically written under /opt/j2se.

Good Luck,
Federico

 
 
 

1. Disk full on non-full disk???

On a Dell v2.2 system I have a 2G disk partitioned as ~300M root and
~1.4G /home (I know it doesn't add up :-)
The system is saying the /home partition is full even though df reports
blocks and inodes available?

$ df -vi
Mount Dir  Filesystem        blocks     used     free %used iused  ifree %iused
/        /dev/root           601562   460352   141210  76%  22896 171664  11%
/proc    /proc                    0        0        0   0%     75    951   7%
/dev/fd  /dev/fd                  0        0        0   0%     66      0 100%
/home    /dev/dsk/0s3       2813640  2439692   373948  86%  92123 975141   8%
/stand   /dev/dsk/0s10        14336    11044     3292  77%      8    128   5%

Could this be some sort of kernel table overflow/corruption, some parameter needs
t
--
Mark Powell


Work:   +44 61 745 3376         Fax:    +44 61 736 3596

2. What's left before Linux 2.0?

3. Disk Full Handling

4. fatal signal 4 ?

5. "full disk", not full

6. can`t start Javacode

7. Hard Disk Weirdness - suddenly disk full

8. Enable the Alt=Meta in XF86

9. disk full while creating emergency disk

10. " File System full " error message : but the partition is not full.

11. filesystem full -> It's not full

12. Q: Handling large mail loads....how to distribute mail handling??

13. /tmp full, swap full?