[Aces-support] trouble with local scratch on ao

[Susan C Dunne]

Hi,
I'm running some code on ao, but the /scratch on some of the nodes is 80-100%
used, so I cannot upload the data files I need or have space to output.

Does local scratch self-clean when it reaches some threshold or is there any way
to tidy it up so these nodes are usable?

Thanks,
Susan

ps:: Here is an example of local scratch usage on some of the nodes:
>> for f in `cat listnodes`; do echo $f `ssh $f df -h /scratch | tail -n1`; done
a54-1727-043 /dev/sda6 17G 2.2G 14G 15% /scratch
a54-1727-044 /dev/sda6 17G 2.3G 14G 15% /scratch
a54-1727-046 /dev/sda6 17G 2.3G 14G 15% /scratch
a54-1727-047 /dev/sda6 17G 12G 3.8G 76% /scratch
a54-1727-048 /dev/sda6 17G 16G 0 100% /scratch
a54-1727-049 /dev/sda6 17G 5.9G 9.4G 39% /scratch
a54-1727-050 /dev/sda6 17G 13G 3.0G 81% /scratch
a54-1727-051 /dev/sda6 17G 13G 2.8G 82% /scratch
a54-1727-052 /dev/sda6 17G 15G 444M 98% /scratch
a54-1727-053 /dev/sda6 17G 14G 2.3G 86% /scratch
a54-1727-055 /dev/sda6 17G 14G 2.3G 86% /scratch
a54-1727-056 /dev/sda6 17G 14G 1.9G 88% /scratch
a54-1727-057 /dev/sda6 17G 4.7G 11G 31% /scratch
a54-1727-058 /dev/sda6 17G 14G 2.1G 87% /scratch
a54-1727-059 /dev/sda6 17G 4.5G 11G 30% /scratch
a54-1727-060 /dev/sda6 17G 12G 3.3G 79% /scratch
a54-1727-061 /dev/sda6 17G 4.5G 11G 30% /scratch
a54-1727-062 /dev/sda6 17G 850M 15G 6% /scratch
a54-1727-063 /dev/sda6 17G 4.5G 11G 30% /scratch
a54-1727-067 /dev/sda6 17G 1.7G 14G 11% /scratch
a54-1727-068 /dev/sda6 17G 2.3G 13G 15% /scratch
a54-1727-073 /dev/sda6 17G 5.5G 9.9G 36% /scratch
a54-1727-074 /dev/sda6 17G 5.6G 9.7G 37% /scratch