4 ms·
If you are relying on failures and crashes instead of proactively monitor your workload you are already taking the wrong approach, swap or no swap.
by goodpoint 2y ago
If you are relying on failures and crashes instead of proactively monitor your workload you are already taking the wrong approach, swap or no swap.
- shermantanktop 2y agoYeah, but shit happens. Oom killer and the load balancer ejecting the host is better than 0.03% of requests mysteriously failing for weeks.