I just bump it to 3GB and it uses all the swap fil...
# fleet
I just bump it to 3GB and it uses all the swap file once the vulnerability cron kicks on lol
Our initial testing indicated it uses ~4GB of memory to run the vuln scan. I wonder if you bump your page file further if it will work?
4GB! Goodness
@zwass the problem with that is that I am using the tmp scratch drive of the Azure VM which is 3.7GB total size, so I could not do that. I could I guess move swap to the main volume but I wouldn't want to do that indefinitely. It may need to be a documented limitation that a fleet server now needs at least 4GB memory to run all the bells and whistles and I write that feature off until I can justify a bigger server. Funny how 1 feature means you need to bump up to a server that's capable of running windows 😅 doesn't seem very efficient, but I say that without knowing how many entries in the NVE database. probably there is a lot and you're building an Array in RAM to compare against the installed softwares.
I will also explore using an ARM A1 instance on OCI, because I have some 12GB RAM instances on there I could use
it's definitely not an easy problem. If we make it consume less RAM, it can take a prohibitively amount of time for some setups. So this was the compromise we went for
if you could create tickets for the other issues you're seeing, we'll look into them as soon as possible
@Tomas Touceda I think the issues are all related to RAM exhaustion anyway