Re: [ldm-users] high memory

Hi, Kevin,

ext4 appears, so far, to be a food fit for our gluster services. That's all we've looked at it for, and for that matter, we're still using XFS on our gluster system for the HPC instance.

XFS has proven sensitive to loss of a RAID shelf, or loss of a particular drive or two on a shelf, in our world when we were using LVM2 to create larger volumes. That can cause a lot of problems all of a sudden. When a hardware failure happens and XFS goes offline, an xfs-recovery operation often requires loss of data when the XFS log is blown away, which you shouldn't HAVE to do, but, in our experience, is required about 2/3d of the time.

That said, if we never had hardware failures, XFS would be just about perfect...

gc

Kevin R. Tyle wrote:
Hi Gerry, et al.:

Any sense as to whether ext4 is improved enough to warrant its use in lieu of a non-journaling fs such as ext2?
Also, what particular problems are of concern with XFS?

--Kevin

______________________________________________________________________
Kevin Tyle, Systems Administrator               **********************
Dept. of Atmospheric & Environmental Sciences   ktyle@xxxxxxxxxxxxxxxx
University at Albany, ES-235                    518-442-4578 (voice)
1400 Washington Avenue                          518-442-5825 (fax)
Albany, NY 12222                                **********************
______________________________________________________________________

On 10/25/2010 08:43 PM, Gerald Creager wrote:
I have not, but with the amount of disk we run, we've gone far away from ext3. We have XFS for our NFS partitions, and ext4 for our Gluster. There are a whole lot of problems associated with ext3, in cluding but not limited to how long it takes to do one of its periodic checks when you've exceed the time limit or reboot count.

gerry

Arthur A. Person wrote:
Has anyone tried mounting ext3 with the data=writeback option (with the risk that active data files could be corrupt after a crash)? I'm wondering how close performance would come to ext2...

                              Art

On Mon, 25 Oct 2010, Gerald Creager wrote:

The choices for files of this size are ext2 if you don't care about journaling, or XFS. XFS has its own share of problems but does work very well.

gerry

Tyler Allison wrote:
We switched off of ext3 for our data partitions for high speed IO. It blows chunks compared to ext2. You loose journaling but I don't care. I don't have long term storage.

Sent from my iPhone

On Oct 25, 2010, at 3:10 PM, "Jeff Lake - Admin" <admin@xxxxxxxxxxxxxxxxxxxx> wrote:

yep both 7200 SATA2's
was considering a Velociraptor 10K RPM but at the xtra $60/m .... need to really consider long and hard


Jeff,
If your disk RPM isn't at least 7200, your I/O is going to be bad with Level2/3 feeds. *******************************************************************************
Gilbert Sebenste ********
(My opinions only!) ******
Staff Meteorologist, Northern Illinois University ****
E-mail: sebenste@xxxxxxxxxxxxxxxxxxxxx ***
web: http://weather.admin.niu.edu ** *******************************************************************************
_______________________________________________
ldm-users mailing list
ldm-users@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/
_______________________________________________
ldm-users mailing list
ldm-users@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/

_______________________________________________
ldm-users mailing list
ldm-users@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/

--
Gerry Creager -- gerry.creager@xxxxxxxx
Texas Mesonet -- AATLT, Texas A&M University
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843

_______________________________________________
ldm-users mailing list
ldm-users@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/

Arthur A. Person
Research Assistant, System Administrator
Penn State Department of Meteorology
email:  person@xxxxxxxxxxxxx, phone:  814-863-1563


_______________________________________________
ldm-users mailing list
ldm-users@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/

--
Gerry Creager -- gerry.creager@xxxxxxxx
Texas Mesonet -- AATLT, Texas A&M University
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843



  • 2010 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the ldm-users archives: