amazon web services - Disk usage when redshift doing vacuum merge? -
i know redshift split vacuum progress 2 stages: sort , merge.
during sorting, disk usage not change, merge stage seems occupy lot of free space.
my cluster 3 nodes of dw2.xlarge, total 480 gb ssd. before vacuuming, total disk usage around 50%.
i'm doing vacuum on a table of 81gb, failed when trying merge, due disk full error.
i want know how space should reserve vacuuming large unsorted table?
i asked question redshift support team, haven't got reply until now. have experience it?
yes longer vacuum use deep copy
this should avoid disk usage problems.
while insert ( select * from...) here in select select sorted order data data inserted in sorted fashion.
do incremental insert (select * ) per sort key e.g if storing data 30 days deep copy day day should avoid space issues
Comments
Post a Comment