The peer comparison figures have been sourced from Morningstar data and is therefore limited to the funds and investment products included in their database. Performance figures for periods greater than one year are annualised and presented as 'per annum' values. brokerage and assuming dividends re-invested and no withdrawals. # Performance figures are after management and admin fees excl. For more information about fees and costs, please see the Product Disclosure Statement and Investment Menu.

Indirect Costs are estimated to be between 0.09% - 0.30% p.a. * InvestSMART’s capped management fees (capped at $550 p.a.) do not include indirect cost ratio charged by ETF providers nor brokerage. You should consider the PDS, FSG and TMD before making a decision about a product. The PDS for financial products offered through InvestSMART can be downloaded from this website or obtained by contacting 1300 880 160. I would appreciate any answer or insight here, but I’m sure a blog post solely on scaling architecture would be really, really helpful to not only me but also the others in the community! If you know any links which go over scaling architecture on Meteor apps, please do share! Thanks.IMPORTANT: This information is general financial product advice only and you should consider the relevant Product Disclosure Statement (PDS), Financial Services Guide (FSG), Target Market Determination (TMD) or seek professional advice before making any investment decision.

What were the specs of your Compose installment? How big was your replica set, etc.?.

Would you have gone with Galaxy or DigitalOcean? If not, do you expect that Galaxy or DigitalOcean would provide the same kind of capabilities in scaling?.

some instances handle main computation, others handle registration, others handle cron jobs, etc.)

Did each droplet represent an identical instance of the app, or did you separate different concerns between each instance (e.g.

How many droplets/instances on modulus did you use?.

Assuming that you scaled using DDP and not GQL/ApolloĬan you guys talk more about the specific architecture you used in the Crossroads “Brave” app to get it to 20k users? If I pinged the wrong people, feel free to ping the right people much appreciated!

Crossroads brave app