Hi,
Just got a smaller Lustre system running with 1 mdt/mgs and 2 ost''s. On
the
osts i''ve added 1x10GB disk and 1x20GB disk
on each system and formatted and mounted them as /mnt/ost1-4 (ost1) and
/mnt/ost2-3 (ost2). I''ve also mounted the lustre system on the mdt/mgs
server with mount -t lustre x.y.n.p at tcp:/testfs /mnt/lustre. So far so good,
however df -h only reports 9/60GB avail where lsf df -h reports 54.4Gavailable.
172.16.x.y at tcp:/testfs
60G 50G 9.0G 85% /mnt/lustre
vs
filesystem summary: 59.1G 4.7G 54.4G 7% /mnt/lustre
--
lustre-client:/mnt/lustre# du -hcs
4.0K .
4.0K total
I cant find anything on this in the lists or manual, but do you have to
have equal size disks to get all the usable space available to the system
mounting lustre? Is this somehow connected to stripe-size or some automatic
failover settings. My original intent was to create a large storage area
using lustre for numerous clients to mount and use as diskspace.
--
Timh Bergstr?m
System Administrator
Diino AB - www.diino.com
:wq
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.lustre.org/pipermail/lustre-discuss/attachments/20080212/5126beb6/attachment-0002.html
Hi again, Discard my e-mail - I compiled to early in the morning and included the --enable-mindf option in configure. After re-reading the configure options and great help from adilger and scjody in #lustre I finally figured out that I was overseeing the whole stripe thing. Best Regards, Timh 2008/2/12, Timh Bergstr?m <timh.bergstrom at diino.net>:> > Hi, > > Just got a smaller Lustre system running with 1 mdt/mgs and 2 ost''s. On > the osts i''ve added 1x10GB disk and 1x20GB disk > on each system and formatted and mounted them as /mnt/ost1-4 (ost1) and > /mnt/ost2-3 (ost2). I''ve also mounted the lustre system on the mdt/mgs > server with mount -t lustre x.y.n.p at tcp:/testfs /mnt/lustre. So far so > good, however df -h only reports 9/60GB avail where lsf df -h reports > 54.4G available. > > 172.16.x.y at tcp:/testfs > 60G 50G 9.0G 85% /mnt/lustre > > vs > > filesystem summary: 59.1G 4.7G 54.4G 7% /mnt/lustre > > -- > > lustre-client:/mnt/lustre# du -hcs > 4.0K . > 4.0K total > > I cant find anything on this in the lists or manual, but do you have to > have equal size disks to get all the usable space available to the system > mounting lustre? Is this somehow connected to stripe-size or some automatic > failover settings. My original intent was to create a large storage area > using lustre for numerous clients to mount and use as diskspace. > > -- > Timh Bergstr?m > System Administrator > Diino AB - www.diino.com > :wq-- Timh Bergstr?m System Administrator Diino AB - www.diino.com :wq -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.lustre.org/pipermail/lustre-discuss/attachments/20080212/e9dcf3ac/attachment-0002.html
The discrepancies between lfs df -h and df -h I think that largely have to do with the way GB are calculated. Ether based on 1000 or 1024. Remember you lose about 5% of each underlying volume due to the 5% reserved for root. Aaron On Feb 12, 2008, at 8:38 AM, Timh Bergstr?m wrote:> Hi, > > Just got a smaller Lustre system running with 1 mdt/mgs and 2 ost''s. > On the osts i''ve added 1x10GB disk and 1x20GB disk > on each system and formatted and mounted them as /mnt/ost1-4 (ost1) > and /mnt/ost2-3 (ost2). I''ve also mounted the lustre system on the > mdt/mgs server with mount -t lustre x.y.n.p at tcp:/testfs /mnt/lustre. > So far so good, however df -h only reports 9/60GB avail where lsf df > -h reports 54.4G available. > > 172.16.x.y at tcp:/testfs > 60G 50G 9.0G 85% /mnt/lustre > > vs > > filesystem summary: 59.1G 4.7G 54.4G 7% /mnt/lustre > > -- > > lustre-client:/mnt/lustre# du -hcs > 4.0K . > 4.0K total > > I cant find anything on this in the lists or manual, but do you > have to have equal size disks to get all the usable space available > to the system mounting lustre? Is this somehow connected to stripe- > size or some automatic failover settings. My original intent was to > create a large storage area using lustre for numerous clients to > mount and use as diskspace. > > -- > Timh Bergstr?m > System Administrator > Diino AB - www.diino.com > :wq _______________________________________________ > Lustre-discuss mailing list > Lustre-discuss at lists.lustre.org > http://lists.lustre.org/mailman/listinfo/lustre-discussAaron Knister Associate Systems Analyst Center for Ocean-Land-Atmosphere Studies (301) 595-7000 aaron at iges.org -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.lustre.org/pipermail/lustre-discuss/attachments/20080212/f41b7db3/attachment-0002.html