[Ocfs2-users] ls

[Zhen Ren]

--
Best regards,
Eric, Ren
HA team, SU


>>> 
> leopoldo, 
>  
> why did you use DRBD to share only one storage? 

+1

As far as I seen, ocfs2 typically runs on  DRBD with Primary/Secondary mode.
By the way, Primary/Primary seems to be used in cases where KVM/Xen vms be hosted.

>  
> >> 
> >> ub-ocfs:~# service drbd status 
> >> drbd driver loaded OK; device status: 
> >> version: 8.3.13 (api:88/proto:86-96) 
> >> srcversion: 697DE8B1973B1D8914F04DB 
> >> m:res      cs         ro               ds                 p  mounted   
> fstype 
> >> 1:wwwdata  Connected  Primary/Primary  UpToDate/UpToDate  C  /opt     ocfs2 
> > 
> > the ds "Primary/Primary" is in question. 
> > 
> > It seems that DRBD  cannot work when both nodes is primary. 
> > 
> > This doc may help you get some hint. 
> >  
> https://www.suse.com/documentation/sle_ha/book_sleha/data/sec_ha_drbd_configu 
> re.html 
> > 
> > Also,inproper configuration might lead to this issue. 
> > 
> >> 
> >> ub-ocfs:~# ls /opt/ 
> >> local  lost+found 
> >> 
> >> ub-ocfs:~# ls /opt/local 
> >> Segmentation fault 
> > 
> > IIRC,I also got seg-fault problem for some unclear reason,which may be  
> caused by setting up problem. 
> > 
> >> 
> >> ub-ocfs:~# ps fa 
> >>   PID TTY      STAT   TIME COMMAND 
> >>  2472 pts/2    Ss     0:00 -bash 
> >>  2529 pts/2    R+     0:00  \_ ps fa 
> >>  2299 pts/1    Ss     0:00 -bash 
> >>  2356 pts/1    S+     0:00  \_ -bash 
> >>  2361 pts/1    D+     0:00      \_ ls -l /opt/local/etc /opt/local/games 
> >> /opt/local/include /opt/local/leo1.... 
> >> 
> >> 
> >> the script ls -l /opt/local/* 
> > 
> > You mean ls -l /opt/local go bad, but ls -l /opt/local/* can work well? 
> > 
> >> 
> >> ub-ocfs:~# cat /opt/local/leo-ocfs6.sh 
> >> date ; echo -n 
> >> hostname 
> >> ls -l 
> >> ls -l /opt/local/le* 
> >> ls -l /opt/local/* 
> >> 
> >> 
> >> I can work normal 
> > 
> > what do you mean by work normal? ;-) 
> > other cmd can work, except ls -l /opt/local, though seg-fault already  
> happened? 
> > 
> >> 
> >> ub-ocfs1:~# rm -r /opt/local/bin 
> >> ub-ocfs1:~# mkdir /opt/local/etc/bin 
> >> ub-ocfs1:~# ls /opt/local/leo* 
> >> /opt/local/leo1  /opt/local/leo3  /opt/local/leo-ocfs    
> /opt/local/leo-ocfs2 
> >> /opt/local/leo-ocfs4  /opt/local/leo-ocfs6 
> >> /opt/local/leo2  /opt/local/leo4  /opt/local/leo-ocfs1   
> /opt/local/leo-ocfs3 
> >> /opt/local/leo-ocfs5  /opt/local/leo-ocfs6.sh 
> >> 
> >> 
> >> but resource is busy if I want restart or unload 
> >> 
> > 
> > don't worry about this. 
> > when the previous problem done, this would go away. 
> > 
> >> ocfs:~# service o2cb unload 
> >> Stopping ocfs2_controld.cman: /sbin/start-stop-daemon: warning: this system  
> is 
> >> not able to track process names 
> >> longer than 15 characters, please use --exec instead of --name. 
> >> Failed 
> >> Unable to unload modules as the cluster is still online 
> >> 
> >> 
> >> because '2361 pts/1    D+     0:00      \_ ls -l /opt/local/' is in  'D 
> >> uninterruptible sleep (usually IO)' 
> >> 
> >> if I want the control back I need reboot host, 
> >> 
> >> anybody have another solution ?-:-( 
> > 
> > hm,what's your solution? ;-) 
> > 
>  
>