I think we can start by migrating to the free DB host, and test everything out locally. If it doesn’t go well, we can look at the other options.
1 Like
What do you think of the issues @tmcnulty expected us to encounter on that path?
Perhaps we should try to fix it on MongoDB anyway as it suggests something isn’t set up right, and they are now offering us technical support to try and get things running. What was the error message you found Tom?
I didn’t see this until now. I will read through to understand better because it seems I must have skipped some messages.
I also want to focus on getting the floating footer working because multitasking is not friendly lol 
1 Like
I think Tom is getting the dumps ready for Jordin while i’m preparing a proposal statement, and that …should have us up and running again
Update 29/4: email from Mongo
It’s still unclear what exactly is causing the problem as I don’t have a full picture of the environment or conditions. That said, the feedback you received earlier does seem to correlate with some of the info you’ve shared.
In particular, the error stack does appear to show that the meteor app is failing due to a listIndexes command being run on the admin database. Also, per the link you sent, the Atlas free (M0) and shared tier (M2 & M5) clusters do not allow read/write access to the admin database. The question, however, is why would the code be running this command on the admin database? This is not typical…
Following up on some of the feedback you’ve received, it’s clear that your code is calling ensureIndex (https://github.com/focallocal/fl-maps/blob/ea6c71a1a1d24440f1be493d88a9c4e681203708/imports/both/collections/events/index.js#L572). From the code that was linked, the ensureIndex function (https://github.com/mongodb/node-mongodb-native/blob/2b18411d2f57e06d11262d5a308c56a9f561789e/lib/operations/db_ops.js#L132) calls indexInformation (https://github.com/mongodb/node-mongodb-native/blob/2b18411d2f57e06d11262d5a308c56a9f561789e/lib/operations/db_ops.js#L279), which in turn calls listIndexes (https://github.com/mongodb/node-mongodb-native/blob/2b18411d2f57e06d11262d5a308c56a9f561789e/lib/operations/db_ops.js#L305) which appears to be where the problem is triggered.
However, as per some of the other feedback you’ve received it may just be that you are inadvertently running this code on the admin database when in fact it should be run on a user-defined database. This could easily be the result of a slight misconfiguration in your code - specifically with the connection string.
To verify - could you please provide the full MongoDB connection string details (Note: please be sure to strip out the actual password if present - this is not required). Looking at your code this is likely the unencrypted version of the mongo_url variable either from https://github.com/focallocal/fl-maps/blob/ea6c71a1a1d24440f1be493d88a9c4e681203708/deployment/brightertomorrowmap/mup-secrets.enc.json#L2 or from https://github.com/focallocal/fl-maps/blob/ea6c71a1a1d24440f1be493d88a9c4e681203708/deployment/publichappinessmovement/mup-secrets.enc.json#L2
This is the same info as you would have received from the ‘connect to your cluster’ dialog in Atlas itself and would typically look similar to the following:
mongodb+srv://user:password@cluster0.mongodb.net/test?w=majority
This may be stored as a single string or split into parts - either way, can you share the details please (minus the password of course)?
Andy et al,
I was able to get the connection strings necessary after getting the SOPS keys updated today as we discussed.
mongodb+srv://USER:PASSWORD@fl-sleeper.mxuvf.mongodb.net/fl-sleeper
mongodb+srv://USER:PASSWORD@fl-gather.mxuvf.mongodb.net/fl-gather
Additionally I have dumps of the databases so we can populate alternate providers as previously discussed. I believe it was @JordinParma that offered to take a look at this?
Dumps via Google Drive. I believe sharing on this item is currently restricted, once someone requests access I’ll confirm the user and then allow the downloads.
Sorry this took so long, looking forward to having my regular development environment back together.
1 Like
Well done Tom. I was expecting something far bigger, but if thats what the guy needs thats awesome!
@tmcnulty @zofrex @SysAdmin Russel asks:
Have you performed a query in mongo atlas to list some documents and look at the data in the documents?
Update!
I strongly believe this issue is resolved thanks to @carlos’s efforts this weekend, and everyone who helped explore the issue in the past. Our platform should go live again tomorrow!!!
Were back!!!
Huge thanks to @tmcnulty @syl @zofrex @carlos and everyone else that joined in to help get us running again!
Lets get this project out and helping as many people as we can again!
The final pre-launch steps are the new homepage, and i’ll be updating the about text, whitepapers, and a few other things. As soon as we cross those fairly small tasks off off then its time to start telling everyone about our platform and community led mission to create a world with more peace and well-being