[aether-announce] BeeGFS system backup running

[Marius Dan]

Dear Aether Users,

The faulty SSDs storing the BeeGFS metadata have been replaced. The data has been successfully copied back on the new SSDs from the much slower HDDs. Thus, everything is back to normal on the BeeGFS storage system. 

Thank you for your patience!

Best wishes,

Marius

> On 14. Sep 2020, at 13:24, Marius Dan <mdan at uni-bremen.de> wrote:
> 
> Dear Aether Users,
> 
> The backup has finished. Please note that the faulty SSDs have been replaced with HDDs and the BeeGFS performance for the next few days will be pretty low. This is because on two out of three BeeGFS servers the metadata is written to HDDs instead of SSDs. Things will return to normal once we receive the warranty SSD replacements. 
> 
> Thank you for your understanding!
> 
> Best wishes,
> 
> Marius
> 
>> On 14. Sep 2020, at 11:38, Marius Dan <mdan at uni-bremen.de> wrote:
>> 
>> Dear Aether Users,
>> 
>> Two SSDs storing the BeeGFS metadata files have failed on the storage servers and we urgently need to run an up-to-date backup of the metadata files. The backup is currently running on all three BeeGFS servers and it could impact your simulations if you are doing intensive I/O. The backup process will take about on hour to finish and it should finish before 13:00. 
>> 
>> Sorry for any inconvenience!
>> 
>> Best wishes,
>> 
>> Marius
>> _______________________________________________
>> aether-announce mailing list
>> aether-announce at mailman.zfn.uni-bremen.de
>> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/aether-announce
> 
> _______________________________________________
> aether-announce mailing list
> aether-announce at mailman.zfn.uni-bremen.de
> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/aether-announce