I have just added a RAID 5 array to my openSUSE 12.1 64 bit box and seek advice on creating a mount point so that I can move my multimedia files to the new drive.
Using Yast>System>Partitioner the drive shows as /dev/sdb and if I select the whole capacity and go to on partition it I am given the option of mounting it but only offers /srv /tmp or /local. My objective is to have my multimedia stuff on this drive as a directory in my home directory. How should I proceed please?
Budgie
Hold on this please. Found a thread which answers it.
On 2012-10-03 22:46, Budgie2 wrote:
> t I am given the
> option of mounting it but only offers /srv /tmp or /local.
Type any name you like.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
Yes, like Carlos. Those two names are only a suggestion. You can use any place in your directory tree you like. Thus also something like /home/budgie2/multimedia
Only thing is you check after the action if /home/budgie/multimedia is owned by user budgie and group users. When not use chown (as root).
Hi Carlos & Henk,
I need some advice here. My system is as follows:-
OS: Linux 3.1.10-1.16-desktop x86_64
System: openSUSE 12.1 (x86_64)
KDE: 4.7.2 (4.7.2) “release 5”.
My first RAID array shows as 465 GB /dev/sda which has three primary partitions sda1 at 22 GB, sda2 at 23 GB and sda3 at 24 GB. The operating system is on the middle one where root is mounted. The other two are intended for when I wish to try other operating systems but are not in use at present. sda4 is an extended partition of 396 GB containing sda5 at 394 GB mounted as /home and sda6 a 2GB swap partition.
I was getting warnings that my drive was nearly full, which is what prompted the acquisition of 4x1Tb drives for a new array. I backed up my multimedia and then deleted all the files and the directory using Dolphin. What a pain that was because of restrictions in size of wastebin and some files in there I couldn’t get rid of even as root from console. Eventually I managed to clear everything out and remove the multimedia directory in readiness to use this as mount point for my new array.
Now I cannot be certain of sequence of events but after I had built the second RAID array but before I formatted it or shortly after I had started the formatting I used System Information to see what was recorded there out of interest. Couldn’t do much else with the new drive which took 16 hours to format. Under “Disk Information” the root partition is shown in red and as being full. What on earth could have filled it and what should I do to clear some space before my machine chokes up?
Budgie2
This is a whole story and I can not grab all of that in such a short time.
A few remarks.
When there is warmning that a disk (partition) is getting full, one does not by a disk, but one first tries to find out what the course is. A loging file that runs wild? A /tmp that is never cleaned? Logical because youare storing 7 movies a day?
Also there is no warning that a “drive” is full. There is a warning that a disk (partition) is full. Btw you see that with
df
And that brings me to another remark. You tell a lot of information that should clarify your computer facts. But there are no computer facts. Talking about partitions starts with
fdisk -l
OK point taken. Here is what I have.
Disk /dev/sda: 500.0 GB, 499961036800 bytes
255 heads, 63 sectors/track, 60783 cylinders, total 976486400 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000b9824
Device Boot Start End Blocks Id System
/dev/sda1 63 46154744 23077341 83 Linux
/dev/sda2 * 46154745 94381874 24113565 83 Linux
/dev/sda3 94381875 144713519 25165822+ 83 Linux
/dev/sda4 144713520 976462829 415874655 f W95 Ext'd (LBA)
/dev/sda5 144713583 972253799 413770108+ 83 Linux
/dev/sda6 972253863 976462829 2104483+ 82 Linux swap / Solaris
WARNING: GPT (GUID Partition Table) detected on '/dev/sdb'! The util fdisk doesn't support GPT. Use GNU Parted.
Disk /dev/sdb: 3000.1 GB, 3000071356416 bytes
255 heads, 63 sectors/track, 364737 cylinders, total 5859514368 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdb4 1 1 0+ ee GPT
No idea where GPT comes from but it must be an artifact of using Yast Partitioner.
Grateful for any further advice in due course.
Many thanks.
Budgie2
On 2012-10-04 19:16, Budgie2 wrote:
>
> Hi Carlos & Henk,
> I need some advice here. My system is as follows:-
> OS: Linux 3.1.10-1.16-desktop x86_64
> System: openSUSE 12.1 (x86_64)
> KDE: 4.7.2 (4.7.2) “release 5”.
>
> My first RAID array shows as 465 GB /dev/sda which has three primary
> partitions sda1 at 22 GB, sda2 at 23 GB and sda3 at 24 GB. The
> operating system is on the middle one where root is mounted. The other
> two are intended for when I wish to try other operating systems but are
> not in use at present. sda4 is an extended partition of 396 GB
> containing sda5 at 394 GB mounted as /home and sda6 a 2GB swap
> partition.
Like Henk, I prefer to see the output of commands showing things, instead of descriptions 
In this case, the output of “fdisk -l” would show the partitions at a glance, and the output of
“mount” would show where they are mounted. Also, “df -h” would show the available space on all
of them.
Use code tags to paste all that.
Posting in Code Tags - A Guide
> I was getting warnings that my drive was nearly full, which is what
> prompted the acquisition of 4x1Tb drives for a new array. I backed up
> my multimedia and then deleted all the files and the directory using
> Dolphin. What a pain that was because of restrictions in size of
> wastebin and some files in there I couldn’t get rid of even as root from
> console.
When you want to move huge amounts of data, do not use graphical tools; use midnight commander
(mc) in a text terminal or virtual terminal (text mode). Doing it in text mode, without you
been logged in kde allows to move those unmovable files because they are in use. Also, mc does
not use a dustbin.
> Eventually I managed to clear everything out and remove the
> multimedia directory in readiness to use this as mount point for my new
> array.
And again, you did not need to remove those files. Instead, you could simply have renamed that
multimedia directory to something else, and that allows you to create the same directory again,
and empty. Later, you can move (or copy, delete later after checking) the files from the old
place to the new without using the backup.
> Now I cannot be certain of sequence of events but after I had built the
> second RAID array but before I formatted it or shortly after I had
> started the formatting I used System Information to see what was
> recorded there out of interest. Couldn’t do much else with the new
> drive which took 16 hours to format.
That would be something else you did, formatting is an instant.
> Under “Disk Information” the root
> partition is shown in red and as being full. What on earth could have
> filled it and what should I do to clear some space before my machine
> chokes up?
Again, you can use midnight commander to show the sizes of directories and find out. Or
graphical tools like baobab (gnome), kdirstat, filelight… or console tools like df and du.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
On 2012-10-04 21:16, Budgie2 wrote:
>> This is what I have now:-
>>>
> Code:
> --------------------
> > > Filesystem 1K-blocks Used Available Use% Mounted on
> > rootfs 23735128 22702068 0 100% /
> > devtmpfs 4605712 36 4605676 1% /dev
> > tmpfs 4613460 80 4613380 1% /dev/shm
> > tmpfs 4613460 560 4612900 1% /run
> > /dev/sda2 23735128 22702068 0 100% /
> > tmpfs 4613460 0 4613460 0% /sys/fs/cgroup
> > tmpfs 4613460 560 4612900 1% /var/lock
> > tmpfs 4613460 560 4612900 1% /var/run
> > tmpfs 4613460 0 4613460 0% /media
> > /dev/sda5 407277444 58863216 327725724 16% /home
> > /dev/sdb1 2883781880 205808 2737088316 1% /home/alastair/mastermedia
> > 192.168..:/Qmultimedia/ 960301840 748507080 211794760 78% /home/alastair/QnapMultiM
> > 192.168..:/Multimedia/ 5809699264 2496833664 3312341312 43% /home/alastair/Qnap2MultiM
> --------------------
your new multimedia space shows there on sdb1, I understand. Home has little use, and your root
filesystem is full - you have to investigate that, I said how on another post.
> OK point taken. Here is what I have.
>
>
> Code:
> --------------------
> WARNING: GPT (GUID Partition Table) detected on ‘/dev/sdb’! The util fdisk doesn’t support GPT. Use GNU Parted.
>
>
> Disk /dev/sdb: 3000.1 GB, 3000071356416 bytes
> 255 heads, 63 sectors/track, 364737 cylinders, total 5859514368 sectors
> Units = sectors of 1 * 512 = 512 bytes
> Sector size (logical/physical): 512 bytes / 512 bytes
> I/O size (minimum/optimal): 512 bytes / 512 bytes
> Disk identifier: 0x00000000
>
> Device Boot Start End Blocks Id System
> /dev/sdb4 1 1 0+ ee GPT
>
> --------------------
>
>
> No idea where GPT comes from but it must be an artifact of using Yast
> Partitioner.
No artifact, it is intentional. You have to use GPT on disks that size. Fdisk can not show them.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
Hi Henk,
I think I have found the culprit. To move multimedia files I used Dolphin and for some reason there were copies of much of the multimedia stuff in the /tmp/kde directory. Deleted that and now I have only 6 GB of root partition used which is fine. Will check backups in due course but please may I ask a couple of supplementary question?
In Dolphin going down the left hand side after Home, Network, Root and Wastebin there are three icons followed by a size ie 394.6 GB, 22 GB and 23 GB which correspond to my /home /root and one of the “spare” partitions which had been formatted. All are labelled as “Removable Media” Since this is very much not the case, why is this?
Going further down the left hand side there are two NFS shares shown and the new RAID drive which is labelled “primary” but has no size or any other detail. Is this as it should be.
Finally, the new drive/partition is shown mounted as mastermedia which is what I intended but although I have not yet done anything other than format the drive there is a folder called “lost+found” and System Information shows 139.9 GiB already used. Is this just an artifact of formatting and what is lost+found all about please?
Thanks again for all your help so far. I am reading up as fast as I can but your prompts and gentle push in the right direction are much appreciated.
Regards,
Budgie2
On 2012-10-04 22:16, Budgie2 wrote:
> Finally, the new drive/partition is shown mounted as mastermedia which
> is what I intended but although I have not yet done anything other than
> format the drive there is a folder called “lost+found” and System
> Information shows 139.9 GiB already used. Is this just an artifact of
> formatting and what is lost+found all about please?
Forget system info, and use other, more reliable, tools.
Df said;
> Filesystem 1K-blocks Used Available Use% Mounted on
....
> /dev/sdb1 2883781880 205808 2737088316 1% /home/alastair/mastermedia
in KiB. So, there are 205808 KiB used, or 200 MiB - not 139 GiB - unless you have done
something else after that df above.
lost+found is where fsck puts files that were lost and it found
If you have a filesystem corruption, fsck tries to reconstruct it al, but sometimes it fails.
Some files get “lost”, ie, the name is not known any more, or the path. Those files are put
into that directory. Do you have something there already?
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
Sometimes it can be useful to use a graphical representation program like Filelight to find out where space is being taken up. It can really speed up locating places where some rogue program is dumping temp or working files.
Hi and thanks for this suggestion. I was wondering if there was such a program. As it was I got lucky with Dolphin and found the offending files although not what caused them to be written. Thanks again.
Budgie2
Hi Carlos,
First thanks for the heads up on mc. It is so much better than Dolphin for what I want.
There is nothing in my lost+found directory on the new drive and I cannot fathom why 200 MiB has been used.
What a pity I didn;t get your post before relying on backup. That would have saved me a whole couple of evenings. At least I shall know next time.
Regards,
Budgie2
A couple more thoughts. As if to emphasise that your method of renaming the directory would have been better my attempts to copy from the backup have thrown up a problem. If I use Dolphin I can use auto-skip on errors but using mc I cannot (do not know how). The two machines are connected by a gigabit lan but even so mc reports 8 hours to do the copy. The trouble is every so often it comes across an error, probably due to my using Dolphin to make the backup and at that point I get the red dialogue box skip, retry, cancel or similar. I came back to my machine by chance to find this and that everything had stopped pending my action. How can I avoid watching the screen for 8 hours? What I would like to do is auto-skip but log the errors so I can ensure master data set is OK. Any ideas?
Second point was that many of the log files I found when trying to sort out my /root partition could not be opened and had ending .xz or no ending at all. There used to be a log viewer in Yast miscellaneous but alas no longer. How do I open these logs please?
Budgie2
BTW is it Robin or Carlos?
On 2012-10-05 21:06, Budgie2 wrote:
>
> Hi and thanks for this suggestion. I was wondering if there was such a
> program.
I told you several. 6, actually.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
On 2012-10-05 23:46, Budgie2 wrote:
>
> A couple more thoughts. As if to emphasise that your method of renaming
> the directory would have been better my attempts to copy from the backup
> have thrown up a problem. If I use Dolphin I can use auto-skip on
> errors but using mc I cannot (do not know how). The two machines are
> connected by a gigabit lan but even so mc reports 8 hours to do the
> copy. The trouble is every so often it comes across an error, probably
> due to my using Dolphin to make the backup and at that point I get the
> red dialogue box skip, retry, cancel or similar. I came back to my
> machine by chance to find this and that everything had stopped pending
> my action. How can I avoid watching the screen for 8 hours? What I
> would like to do is auto-skip but log the errors so I can ensure master
> data set is OK. Any ideas?
No, that’s the only complain I have with mc. For that kind of thing I use rsync instead.
rsync --archive --acls --xattrs --hard-links --stats --human-readable SOURCE/ DEST
> Second point was that many of the log files I found when trying to sort
> out my /root partition could not be opened and had ending .xz or no
> ending at all.
xz is a compressed file, see the wikipedia. This is something you did, not openSUSE.
> BTW is it Robin or Carlos?
Both, but I prefer Carlos. The forum software uses the login name instead of the name.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
Hi Carlos,
At the time I missed the significance of the advice as I was lucky and had found part of the problem already.
Please forgive.
[QUOTE=robin_listas;2493534]On 2012-10-05 23:46, Budgie2 wrote:
>
> A couple more thoughts. As if to emphasise that your method of renaming
> the directory would have been better my attempts to copy from the backup
> have thrown up a problem. If I use Dolphin I can use auto-skip on
> errors but using mc I cannot (do not know how). The two machines are
> connected by a gigabit lan but even so mc reports 8 hours to do the
> copy. The trouble is every so often it comes across an error, probably
> due to my using Dolphin to make the backup and at that point I get the
> red dialogue box skip, retry, cancel or similar. I came back to my
> machine by chance to find this and that everything had stopped pending
> my action. How can I avoid watching the screen for 8 hours? What I
> would like to do is auto-skip but log the errors so I can ensure master
> data set is OK. Any ideas?
No, that’s the only complain I have with mc. For that kind of thing I use rsync instead.
rsync --archive --acls --xattrs --hard-links --stats --human-readable SOURCE/ DEST
I remember mc from early OS/2 days but didn’t realise there was a port for Linux. I have been trying to learn about rsync which is very powerful but rather scary. I am using a gui front end called LuckyBackup at the moment as it helps me learn with safeguards.
Thanks again for your advice.
> Second point was that many of the log files I found when trying to sort
> out my /root partition could not be opened and had ending .xz or no
> ending at all.
xz is a compressed file, see the wikipedia. This is something you did, not openSUSE.
If I did I was not aware and did not do it intentionally.
> BTW is it Robin or Carlos?
Both, but I prefer Carlos. The forum software uses the login name instead of the name.
OK Carlos, I am just happy to have your help.
Regards,
Budgie2
–
Cheers / Saludos,
Carlos E. R.
On 2012-10-06 22:56, Budgie2 wrote:
>
> Hi Carlos,
> At the time I missed the significance of the advice as I was lucky and
> had found part of the problem already.
> Please forgive.
No problem
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)