[theallan]

Should one of the first things you do with a database not be to have a backup strategy? I understand that HA would be a "nice to have" when first starting out, but surly if you have a production db, a backup and restore plan should be on a survival guide? Neither appear to be mentioned here.

What do you all use for your pg backups? Is Barman ( https://pgbarman.org/) still the way many do it? (I haven't deployed a new pg instance for a while, but thinking about it for a new project).

[mjr00]

I might get flak for saying this but if you aren't a postgres expert already: just use RDS or a similar cloud DB. The amount of money you're saving by hosting and managing your own postgres instance is absolute peanuts compared to having battle-tested infrastructure for HA, backup and restores, point-in-time recovery, read replicas, etc.

[dwedge]

At $dayjob we have the same mentality and as a result have a load of managed read replicas that are never used for anything (not reporting, not read only queries, not backups because $cloud handles it) that cost every month. Plus managed database restricts what you can do with the database - sometimes in really annoying ways.

So while I partly agree with you, a lot of companies don't really need HA, read replicas, or even PITR (though I would argue the last one is so trivial and cheap to enable that why not), but they click the expensive check box, and I would argue that companies who do need these features should consider hiring at least a couple of DBAs and get more flexibility instead of the current status quo of everyone being scared of the database and everyone just hoping cloud support will come to their rescue if ever needed

[drdexebtjl]

> hiring at least a couple of DBAs and get more flexibility

Every place I’ve ever worked at that had DBAs had the complete opposite of more flexibility. You have to do things the DBA’s way, and if their way doesn’t work for your service, you need to fight for their time and priority.

Meanwhile every place I worked at where every team completely owned their databases + did periodic data recovery drills had much more flexibility and no data loss.

[sgarland]

That’s usually because they’ve seen a lot of failure modes over the years, and what seems fine to you can have surprising outcomes later.

My personal experience has been the opposite of yours: lots of data inconsistency issues, data loss only resolved by the database team spinning up backups, etc. And somehow, even after yet another incident, there’s never been appetite to properly fix things.

[skullone]

Now you have lots of DBs all probably operating inefficiently with only periodic recovery testing. This is the stuff DBAs do every day, and you're "hasn't happened to us... yet". You'll be just another lesson someday

[drdexebtjl]

Are you saying DBAs would be doing _continuous_ recovery testing? That’s not even a thing that exists.

Managed databases don’t lose customer data. They would be sued and lose, and if word got out, they would lose their customers. They are optimized for the machines they run on. The proprietary tooling from AWS for managing HA clusters, doing blue-green upgrades and monitoring for slow queries is all miles ahead of the open source tooling available for self-managed databases.

There’s nothing left for the DBA to do.

[skullone]

DBAs in orgs who resource their data teams indeed do constant validation of recovery. I know many DBAs who are in more demand than ever due to the rise of managed databases. If you don't need a DBA, you aren't at a scale or maturity just yet. That you're so dismissive of this role, I have to dismiss your entire opinion because you simply don't grasp the scale of large databases

[dwedge]

> There’s nothing left for the DBA to do.

And yet here we still are

[throwaway894345]

> At $dayjob we have the same mentality and as a result have a load of managed read replicas that are never used for anything (not reporting, not read only queries, not backups because $cloud handles it) that cost every month.

Obviously "let RDS manage your database" doesn't require egregious read replicas. The decision to use read replicas or not is completely orthogonal to whether you use RDS to manage them.

[dwedge]

> Obviously "let RDS manage your database" doesn't require egregious read replicas

Of course not, but an easy checkbox, a best practice AWS or terraform guide and someone doing AWS certified X associate makes it easier to happen without anyone ever really discussing it.

> The decision to use read replicas or not is completely orthogonal to whether you use RDS to manage them.

Assuming you're talking about letting RDS manage anything, then sure - apart from it being more likely to slip through the net if nobody has to configure them. Database is just an expensive cost nobody necessarily drills into.

However if you mean the decision to let $cloud manage the replicas (and keep the primary managed), that totally depends on the cloud and the options. For example have you ever tried having a primary in GCP Cloud SQL but the replica not in cloud SQL?

[throwaway894345]

> Of course not, but an easy checkbox, a best practice AWS or terraform guide and someone doing AWS certified X associate makes it easier to happen without anyone ever really discussing it.

I'm defending your employer or their mindset, I'm just disagreeing with your claim that this follows from the parent's cost analysis claims. It _seems_ like your organization's problems are precisely because they _weren't_ doing the kind of cost analysis that the parent advocated. In other words, nothing in the parent's comment advocated for blindly following some Terraform guide. It feels unfair to the parent to suggest that their mindset caused your organization problems when it seems like your organization's problems were caused by _not having_ the parent's mindset.

[ktaraszk]

They won't let you. It's part of their business to keep you locked in.

[throwaway894345]

More likely it’s just not worth going out of their way to support niche deployments like that. I’ve been using various clouds for years, I’ve done half a dozen cloud migrations, currently working at a multicloud org—the vendors aren’t doing much to lock us in. They very much enable us to move platforms by offering things like bulk data transfer tools, standard application runtimes like Kubernetes, workload identity federation (use external identities as principals in the cloud provider’s IAM system, etc).

Like I have no doubt that they’re all greedy bastards, but they aren’t doing much to lock people in. The people who complain about lock in are usually talking about “cloud providers making their services so much easier to use than bespoke platforms on bare Linux hosts such that no one will want to go back to the latter”.

[zdc1]

You can use RDS without HA. Many people probably should do this, but no one wants to tell their boss they want to disable RDS HA and wear 1h of downtime a year so they can cut 50% of their spend.

[williamdclt]

> no one wants to tell their boss they want to disable RDS HA and wear 1h of downtime a year so they can cut 50% of their spend.

If you're a SaaS, 1h of downtime costs a lot of money. In reputation, in lost business, in support time, in on-call response, in making a public postmortem and sharing it to customers and dealing with responses... For saving a few thousand bucks, it's often indeed a very bad deal.

[dwedge]

Most of those are already priced in (support costs, time dealing with a postmortem), and an hour of downtime a year isn't going to meaningfully harm the reputation of most companies, I'd wager a lot of clients wouldn't even notice.

Of course this isn't true for some companies and the cost of downtime far exceeds the cost of HA. Most companies think they are that company, most aren't. And even if they are, "AWS had an outage" does a lot of lifting.

Every time I get a new client they need constant point in time backups, failover, HA everything. Usually when the costs are explained they change their mind and an hour old snapshot restore is actually fine

[h2aichat]

That's really helpful. Thanks!

[vanviegen]

And then.. you're basically trapped inside the AWS cloud (due to egress costs and db latency). No thanks!

[manphone]

There’s a bunch of features in compatibility, performance, and cost reasons not to choose these cloud hosted managed databases.

[inigyou]

This only works if you're already on the cloud, otherwise you're paying through the nose for bandwidth.

[frollogaston]

RDS is nice even if you just want basic functionality with backups.

[9dev]

It’s really not that hard, especially with an AI agent to help you. Running a Postgres server and a read replica in Hetzner with pgBackRest backing up to their S3 buckets and a cloud volume can be had for under 50€, has HA, PITR, 3-2-1 backups, and will carry you through your series A comfortably, with GDPR compliance built-in.

What does the equivalent RDS setup cost you?

[OrangeDelonge]

RDS gets you easy integration with the rest of the AWS services, which as a startup, you’re probably using quite a few of.

Also, if you are nickel and diming over expenses in the 50€ range, you’re probably at the wrong startup.

[9dev]

That was not quite the point I was making, which is that the equivalent RDS setup would come in at around ten times of that.

[williamdclt]

ten times 50euros? Sounds like a very reasonable deal to not have to think about these things or spend the time setting it up.

[9dev]

If expending 6k per year on your database alone sounds reasonable to you, we have very different perspectives on a good use of investor funds.

We’re not talking about maintaining your own k8s cluster here mind you, but a basic Postgres setup. If you can’t handle that comfortably in an afternoon, you probably shouldn’t be entrusted with customer data.

[rsyring]

FWIW, we use: https://pgbackrest.org/

Offers point-in-time recovery which is an improvement over a custom solution we used to have which gave us nightly backups.

We have it backing up to Backblaze B2 (S3 like). Was relatively easy to setup and no problems really.

[CodesInChaos]

There was some recent uncertainty about pgBackRest getting discontinued due to lack of funding. But the maintainer secured funding, and pgBackRest development will continue.

https://pgbackrest.org/news.html

[Tostino]

Can't recommend pgbackrest enough. It's fantastic software, and I love the work they put into doing inter-file deltas for backups (so if 8kb of a 1gb file changes, you only backup the difference). It saved my last company a ton of money on storage while keeping good RTO/RPO.

[k_bx]

It's great, but it's not incremental like git, e.g. you do need to make a full backup periodically, unfortunately.

I didn't understand that, and after 5 months of usage caught my backblaze to be using 40TB, and nightly restores taking forever for other reasons. So: not ideal, and be careful to check!

[Tostino]

Heh, yeah I suppose there are still some foot guns if you don't understand how things work.

[ComputerGuru]

There’s no need to get all complicated and fancy or introduce more dependencies. For most people, a cron job calling pg_dump_all piped to zstd and copying the output to s3/ftp/whatever is plenty good enough.

Obviously past a certain point carting around full backups becomes time/dollar prohibitive, but this can take you very far.

[rsyring]

FWIW, we started with a system that was essentially this. We eventually moved to pgbackrest and it wasn't any harder to setup. But the ROI on that investment is a lot higher because pgbackrest does a lot more for us than the home rolled solution.

Having done both, I'd recommend just starting with pgbackrest.

[Scarbutt]

If you can afford to lose the data created between backups, sure.

[lobo_tuerto]

Better than losing all the data created between no backups.

[Tostino]

But the other option is just doing it right from the start and using a tool like pgbackrest. It's no harder to setup, and it puts you into best practices by default rather than having to work at it later.

I just don't understand why people seem so drawn to the bad solution just because it ships with the database.

[saltcured]

I think you need to analyze the disaster scenarios and recovery costs you are trying to balance before you can really evaluate these different options.

If you are on reasonable storage, I think it is arguable that the main reason you would have to recover is due to a software failure leading to corruption of the PostgreSQL backing state files. Otherwise, you would simply be restarting PostgreSQL on your durable and available storage. So, if the content has been corrupted by bugs, can you trust the recent WAL log in this scenario? Or can a plain dump be a more reliable "known-good" database recovery point?

And what is the business cost to rolling back to a less frequent dump such as nightly? Not every DB is some kind of multi-party OLTP ledger. Sometimes, a day of lost "new data" may be just a day of labor to repeat some data entry or other repeatable work...

A really robust contingency plan probably has both of these kinds of backup, and scenarios where one recovery versus the other is attempted. But if you are starting at a very small scale, hosted on some reasonable cloud storage like EBS, a cron-based dump that goes into a normal hierarchical file backup system may be totally sufficient for the extreme disaster scenario where you cannot simply restart your DB server on top of the surviving and available storage volume?

[Tostino]

Using something like EBS for your postgres data directory is, IMO, a bad idea. If you've already made that mistake, yeah I suppose using PG dump isn't that bad.

I'd much rather just do it right though.

Direct attached storage (or whatever storage is fastest / lowest latency for the environment you have available).

Setup pgbackreset or barman, use WAL archiving and setup block incremental backups with a new full backup every so often.

Now you have point in time recovery with very low RPO/RTO, and your database infrastructure can be treated a bit more like cattle instead of a pet, as you should be testing restores often, and this will become part of your yearly major postgres upgrade strategy.

You also get the ability to have a replica for almost free, as replicas can be created from the backup repo very cheaply, and get caught up to the primary without requiring the primary keep around the WAL for a long time. It just pulls it from the repo until caught up.

Any case where you think EBS was the right choice is better handled by having one (or more) replicas and a failover strategy.

Just do the right thing from the start and you save a lot of headaches. People have run production systems already, and have hit the pain points. Why keep hitting the same ones?

[dwedge]

To play devils advocate, something that doesn't ship with the database is harder to setup than something that does

[CodesInChaos]

An atomic volume snapshot should work for any database that's durable on power failure. Ideally preceded by a checkpoint, to minimize recovery time. Atomicity of the snapshot mechanism is essential to prevent data corruption using this approach.

We used EBS snapshots on AWS for multi-TB MongoDB to get incremental backups that are fast to create and fast to restore (with some performance degradation after restore).

It doesn't support point-in-time recovery, but since it's fast you can create frequent snapshots (e.g. hourly). I'd consider adding this as a secondary backup strategy, even if you use a higher-level postgres-specific backup tool.

[geoka9]

If you already run k8s, why not just use cnpg?

https://github.com/cloudnative-pg/cloudnative-pg

[hoppp]

Backups are mandatory for any serious deployment. But it's more devops and the guide is more about SQL layer.

This guide is only satisfactory if the database is managed, otherwise there are a whole bunch of things going on.

[nickjj]

I go with a simple pg_dumpall approach running on a cron job. It has worked well for over a decade on all sorts of different systems.

Here's a complete walkthrough on how I backup and restore: https://nickjanetakis.com/blog/how-to-back-up-postgresql-in-...

It covers using Plakar too (optionally) if you want deduplication and encryption.

[pphysch]

pgdump / pgrestore, using native binary format