Possible memory leak?

[bitlewis]

We have been using this stratum for a few days now as a test. I've put 228 miners (Antminer S9's) on a single instance and so far it works flawlessly. I also did 2 other instances (total of 3, on 3 different virtual machines) that perform good too.

However, there seems to be a possible memory leak issue.
An instance with 228 miners running has the system at about 780MB of ram during the first few hours.
The instance now has over 48 hours of uptime but the system is consuming 1280MB of ram. Before taking ram measurements I make sure to empty system caches and buffers to get a relevant result.

Would you know what could be causing this and how could we get rid of the increase? Since I planned on running about 70 instances of this proxy Id like to pinpoint what is causing the ram usage to increase before launching it at scale.

Thanks

[fship7]

@louiseth1 did you ever solve this issue? Do you know what causes it? Perhaps related to the database?