| 1 |
Yes, I have a N330 (zotec ION) with 3G ram, no local storage and swap |
| 2 |
over nbd with portage and build area in /tmp which itself is on tmpfs. |
| 3 |
|
| 4 |
Some packages (gcc and glibc in particular) require a lot of ram and |
| 5 |
tmpfs to emerge so sometimes I have to disable tmpfs and use nfs |
| 6 |
storage. |
| 7 |
|
| 8 |
Because portage is lost on reboot when using tmpfs, an emerge sync is |
| 9 |
needed to rebuild it when an update is needed and that takes time (more |
| 10 |
than just updating portage over nfs in fact). The upside is emerges can |
| 11 |
be very fast indeed and updating portage is an unattended operation so |
| 12 |
cost (admin time) is small for me. |
| 13 |
|
| 14 |
Only disadvantage is gcc and possibly gcc need manual intervention to |
| 15 |
disable tmpfs but thats not often - if you dont mind the wasted disk |
| 16 |
space (portage) you can avoid the sync time, but having the build |
| 17 |
directories in tmpfs is a real gain - wish I had the max of 4G though :) |
| 18 |
|
| 19 |
BillK |
| 20 |
|
| 21 |
myth2 linux # df -h |
| 22 |
Filesystem Size Used Avail Use% Mounted on |
| 23 |
devtmpfs 878M 72K 878M 1% /dev |
| 24 |
shm 878M 0 878M 0% /dev/shm |
| 25 |
tmpfs 1.8G 492M 1.3G 29% /tmp |
| 26 |
tmpfs 10M 0 10M 0% /var/lock |
| 27 |
tmpfs 10M 64K 10M 1% /var/run |
| 28 |
tmpfs 10M 320K 9.7M 4% /var/cache/hald |
| 29 |
svcdir 2.0M 208K 1.8M 11% /var/lib/init.d |
| 30 |
myth1:/home/MythTV/videos |
| 31 |
1.2T 960G 251G 80% /mnt/videos |
| 32 |
myth1:/home/MythTV/posters |
| 33 |
1.2T 960G 251G 80% /mnt/posters |
| 34 |
myth1:/home/MythTV/recordings |
| 35 |
1.2T 960G 251G 80% /mnt/recordings |
| 36 |
myth1:/home/MythTV/music |
| 37 |
1.2T 960G 251G 80% /mnt/music |
| 38 |
myth1:/home/MythTV/gallery |
| 39 |
1.2T 960G 251G 80% /mnt/gallery |
| 40 |
myth2 linux # swapon -s |
| 41 |
Filename Type Size Used |
| 42 |
Priority |
| 43 |
/dev/nbd0 partition 2097148 0 |
| 44 |
-1 |
| 45 |
myth2 linux # |
| 46 |
|
| 47 |
|
| 48 |
|
| 49 |
On Wed, 2010-12-15 at 10:15 +0100, Thomas Drueke wrote: |
| 50 |
> Hi, |
| 51 |
> |
| 52 |
> is it possible to emerge packages to a $ROOT directory mounted via NFS ? |
| 53 |
> |
| 54 |
> The setup is |
| 55 |
> - machine A is equipped with a Quad core CPU |
| 56 |
> - machine B is equipped with an N330 Atom-CPU |
| 57 |
> - machine A is doing the system update on a local chroot-environment |
| 58 |
> for machine B and generates binary packages. These packages are |
| 59 |
> installed on machine B using the binary package feature of portage. |
| 60 |
> |
| 61 |
> I expected that the above setup would give an performance improvement |
| 62 |
> over letting machine B do the portage update itself. However a trial run |
| 63 |
> did not show significant improvement that justifies the effort. Machine |
| 64 |
> B still needs a reasonable amount of time to fetch unpack and install |
| 65 |
> the packages. |
| 66 |
> |
| 67 |
> An alternative way might be to mount machine B's / directory via NFS |
| 68 |
> and change make.conf's $ROOT variable to that mount point. |
| 69 |
> |
| 70 |
> Does that sound as a reasonable approach ? |
| 71 |
> |
| 72 |
> Regards, |
| 73 |
> Thomas |
| 74 |
> |
| 75 |
|
| 76 |
-- |
| 77 |
William Kenworthy <billk@×××××××××.au> |
| 78 |
Home in Perth! |
Archives