Remember we are unmounting EOS from the LPC workers tomorrow, but if you have already migrated your scripts to follow the docs below (i.e. using xrootd to access) you should not notice the change.

Thanks!

On Jan 22, 2016, at 11:36 AM, David A Mason <[log in to unmask]> wrote:

> Just a reminder — we will be unmounting EOS from the workers next week.  The actual unmount will happen on _Tuesday_ the 26th (yes _26th_ a day later than originally announced), but please plan to have your LPC workflows updated sooner, so you no longer rely on the mounted filesystems in your condor jobs.  See the link below to instructions how to do interact with EOS properly, and don’t hesitate to ask questions if you have them!
> 
> Thanks,
> 
> —Dave
> 
> On Jan 19, 2016, at 1:13 PM, David A Mason <[log in to unmask]> wrote:
> 
>> As many of you have encountered we have had performance problems with the EOS filesystems on the LPC with the increased Run 2 analysis activity.  To help mitigate this, especially with approvals for winter conferences upon us, we will need to unmount EOS from the LPC workers on January 25.  We understand this is inconvenient, but have worked to make the change as straightforward as possible.  Access to your EOS data from LPC jobs will still be completely possible throughout and how to continue to do that is documented here:
>> 
>> http://uscms.org/uscms_at_work/computing/LPC/usingEOSAtLPC.shtml
>> 
>> Please quickly over this next week adapt your scripts and code to access EOS data via xrootd in advance of the change.  Doing this will work now and continue after EOS is unmounted from the batch farm.  For example within CMSSW use "root://cmseos.fnal.gov/store/user/…"  rather than "file://eos/uscms/store/user/…”
>> 
>> See the link above for more details and how to handle other uses.  Please follow up questions to the lpc-howto list, and we will do our best to keep you productive as you make this change!   
>> 
>> Thanks!   
>> 
>> —Dave