Re: Improving cephfs file listing speed

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi,
  Thank you for the input. Is there any limits on the directories that can be created in the pool. 

Kind Regards
Rabi


From: Sage Weil <sage@xxxxxxxxxxx>
To: Rabindranath Vaz <rabbiv@xxxxxxxxxxx>
Cc: "ceph-users@xxxxxxxxxxxxxx" <ceph-users@xxxxxxxxxxxxxx>
Sent: Wednesday, 17 April 2013, 5:45
Subject: Re: Improving cephfs file listing speed

On Wed, 17 Apr 2013, Rabindranath Vaz wrote:
> Hi,
>    My setup uses ceph bobtail with 3 Metadata, 3 Montors, 6 Storage Nodes.
> The file types stored  are image files of 5MB each, total amount of files
> expected is 5.5 million. As the application is not yet ported to using
> librados currently using cephfs. The ceph kernel module had issues mounting
> and would not start on reboot. The ceph-fuse has been working reliably. File
> saving and retrieving from the app works fine. Listing files  is slow due to
> the number of files. Is there anything i can do to improve the file listing
> performance. When porting to librados can i utilize the files currently
> stored or would I need to backup and restore.
>    Any inputs are much appreciated.

How many files are you storing in each directory?  Keeping the directory
sizes smaller/smallish (1k to 10k files) should help.

You might also try 'mds frag = true' to enable directory fragmentation,
although be aware that this will be less stable.

sage

_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux