This discussion on Hacker News heavily revolves around Tarsnap, primarily in comparison to alternative backup solutions, with a strong emphasis on cost, features, and perceived reliability.
The Dominant Theme: Tarsnap's High Cost Compared to Alternatives
The most frequently raised and strongly held opinion is that Tarsnap is prohibitively expensive, especially when compared to modern alternatives. Users cite significant cost differences, often in orders of magnitude.
- bccdee directly contrasts Tarsnap's pricing with other options: "Storing one terabyte of data in tarsnap costs $250 per month. ... I pay $6 per terabyte per month using backblaze's cloud object storage."
- xnx succinctly captures this sentiment: "50x more expensive than a hard drive feels like a lot."
- homebrewer reiterates this point: "Say rsync.net or borgbase, which are at least 10× cheaper than tarsnap last time I compared them..."
- amar0c stated simply: "It can be whatever it wants I am not paying $25 to store 100GB. I used to use Tarsnap a decade or so ago but pricing makes no sense at all nowadays."
- placardloop is also critical of the pricing: "The pricing isn’t due to AWS. Even if you used standard S3 and paid for data retrieval for your entire backup every single month, tarsnap is over 3x the price of just using S3 yourself. The markup on tarsnap is wild."
- ghostly_s calculates a stark figure: "...three thousand USD a year for 1Tb of cold storage??"
- rafram provides a direct comparison: "OP's cost estimator tells me it would cost a cool $250 per month to keep a terabyte of data backed up in Tarsnap. The same amount costs me $8.25 per month with Backblaze."
- mmh0000 also finds it expensive: "Tarsnap seems very expensive. I'm backing up about 8TiB of data nightly using BorgBackup[0] + InterServer[1] and pay $240/yr."
Restic and BorgBackup as Viable, Cheaper Alternatives
A substantial portion of the discussion highlights restic and borgbackup as powerful, feature-rich, and significantly more cost-effective alternatives to Tarsnap. Many users indicate they have migrated or are considering migrating to these tools.
- snowe2010 initiates the comparison: "Does anyone know how it compares to restic or duplicate?"
- margalabargala provides an initial comparison: "Basically the same service, but much more expensive."
- dividuum states: "restic is basically identical and you can choose where you store your data."
- placardloop elaborates on the flexibility: "That’s because restic is not opinionated about where and how you store your backups. Restic provides a nice interface to create the backups, and then lets you choose where you want to store them (and how access to them is managed), be it locally or via SFTP or S3 or many other backends."
- homebrewer praises these alternatives: "Say rsync.net or borgbase, which are at least 10× cheaper than tarsnap last time I compared them, and can be used with restic or borg which are much faster at restoring even relatively small amounts data..."
- muppetman explains their migration rationale: "I used tarsnap for years, but as my data got bigger and I really wanted to have multipe offsite backups with different providers, I moved to restic. I loved tarsnap - it's a great product. But restic feels very similar but you can backup to your local HD, a remote HD, or "the cloud" and everything is the same CLI commands."
- porridgeraisin gives a strong endorsement: "+1 for restic. Restic + rclone is a very nice combo. Works really well."
- adipid mentions their preferred setup: "I've been considring Hetzner's Storage Box, as it's cheap and I could use just about anything to backup my stuff – although I prefer restic."
- mmh0000 details their successful use of BorgBackup: "I'm backing up about 8TiB of data nightly using BorgBackup[0] + InterServer[1] and pay $240/yr. This gives me differential encrypted rotating backups that are 100% mine and do not lock me into any specific storage vendor."
Tarsnap's Security and Convenience Niche (for a select few)
Despite the overwhelming criticism of its pricing, a few users acknowledge and even appreciate Tarsnap's specific security model and perceived convenience for certain niche use cases, particularly for those who are highly security-conscious or have specific bootstrapping requirements.
- amluto highlights Tarsnap's security advantage: "Tarsnap has a nice security model, and it’s quite a challenge to convince any open-source tool to match it."
- amluto further differentiates Tarsnap's security features: "Tarsnap, in contrast, has an explicit first-class ability to prevent a compromised client from damaging old backups."
- hiAndrewQuinn elucidates a particular niche: "Tarsnap fills but one niche in my overall system. It's a very important niche for which I haven't found any other providers who do anything similar (keyfiles, prepaid, borderline anonymous etc), but it's not where I store the vast majority of my stuff."
- hiAndrewQuinn explains a specific use case related to bootstrapping password managers: "One use case: I don't like the idea of having any accounts at all which I log into without the aid of a password manager. That creates a bootstrapping problem - how am I supposed to log into Google Drive to get my Google Drive password? A prepaid keyfile-based model is one particularly robust way of solving this."
- hiAndrewQuinn also touches on the long-term survival aspect: "That very fact makes it much more likely that Tarsnap actually will survive for those 50 years, which should make us more likely to trust it..." They also mention the prepaid, keyfile-based nature allows for minimal interaction.
- rafram struggles to understand this niche: "I just don't really understand what the niche is. If you have a tiny bit of data that you want to keep backed up and rarely access, you can encrypt it with any number of easy command-line or GUI tools and upload it to Google Drive, Dropbox, or anywhere else with a free tier. If it's securely encrypted, there's no reason to care that the storage provider knows who you are. Tarsnap definitely has nerd appeal, but I can't think of a real problem that it actually solves."
- hiAndrewQuinn responds to this by framing it as a hobby and a preference for extreme paranoia: "I readily admit I'm a nerd about this stuff, and this is primarily a hobby of mine. I am explicitly not 80/20'ing this because it's fun." and further, "If something goes so wrong that my account actually runs out of money before I notice, then I far prefer the default to be "intruder alert, intruder alert, wipe everything". There's a reason it's marketed as backups for the truly paranoid."
Concerns about Reliability and Restore Performance
Beyond cost, Tarsnap's restore speed and the perceived lack of transparency or robust disaster recovery in some alternative services (like BorgBase) were also raised as points of concern.
- homebrewer contrasts restore speeds: "...borg which are much faster at restoring even relatively small amounts data (forget if we're talking terabytes, it's "weeks" vs "your link speed")."
- avian details issues with Tarsnap's monitoring and restore speeds:
- "There's no way to monitor your monthly spend per host/credit left on the account/etc. apart of logging into your account in a browser and manually keeping a spreadsheet."
- "tarsnap restores are slow. Really really slow. A full restore can take days if you have non-trivial amounts of data..."
- "My understanding is that throughput is directly related to your latency to the AWS datacenter where tarsnap is hosted. Outside of north America you can be looking at nearly dial-up speeds even on a gigabit link."
- luizfelberti echoed the slow restore issue: "I also switched away from Tarsnap because I needed to restore my personal PDF collection of like 20GB once and my throughput was like 100Kb/s, maybe less. It has been a problem for at least a decade, with no fix in sight."
- LiamPowell expressed concern about BorgBase's reliability: "Borgbase had a week long (IIRC) outage due to a failed attempt to add new drives to an array. As far as I know they never published a post-mortem on this and have never discussed how they're going to improve their disaster recovery so it can't happen again. It's difficult to recommend when they could leave you without working backups for an entire week."
- avian discusses the trade-offs with Hetzner Storage Box and Borg: "In theory it is less reliable than tarsnap (AWS S3 compared to a single copy on a Hetzner's drive)." and "Borg requires more work to setup and configure compared to tarsnap."
Difficulty in Estimating Tarsnap Costs
Another practical concern raised is the difficulty in accurately predicting Tarsnap's costs due to its deduplication and compression, making it hard for users to budget.
- bigstrat2003 explains this issue: "Ultimately my problem was that there's no way for me to gauge how much the service will cost me. Going just by the amount of data in my home dir, it would be cost prohibitive to upload to Tarsnap. The site does assure me that thanks to compression and deduplication, the actual cost will be far less than I might estimate, which is great! But also, as far as I can tell there's no way to have the client give me an estimate of 'here's how much data you actually have once the secret sauce is applied'."
Manual and Peer-to-Peer Backup Solutions
Some users advocate for more DIY or peer-to-peer approaches as even more cost-effective and control-oriented alternatives.
- pessimizer suggests a collaborative approach: "Also you can back up to the hard drive under your friend's bed, and they can back up to the hard drive under your bed."
- bigstrat2003 confirms this method: "That's what I do! I have a couple of friends from college and we back up to each other over a VPN. It's a very nice solution to the off-site backup need."
- kerblang offers a simple, low-tech solution: "gzip + ccrypt -> thumb drive. Also cozy if your data fits."
- hiAndrewQuinn pushes back on the thumb drive idea with a longevity argument: "I love thumb drives, but Tarsnap is cheaper than the expected 10 year lifetime of a fresh and well maintained thumb drive for the kind of data I hold in there by about a factor of 20 (50 cents vs $10)."
Tarsnap's Pricing Model and its Founder
The pricing model itself, a prepaid system based on actual usage, is recognized but its justification is debated. Some commenters also alluded to the service's founder and his potential motivations or past issues.
- manbash quotes Tarsnap's pricing: "> Tarsnap uses a prepaid model based on actual usage: Storage: 250 picodollars / byte-month of encoded data ($0.25 / GB-month). Bandwidth: 250 picodollars / byte of encoded data ($0.25 / GB)."
- hiAndrewQuinn defends the pricing model in the context of long-term survival and "float": "The business won't survive that long." I'm not so sure. Its ongoing costs appear minimal, and it generates eye watering amounts of float. $5 paid today is >$200 fifty years from now when compounded at 8% real interest. That very fact makes it much more likely that Tarsnap actually will survive for those 50 years..."
- rafram and hiAndrewQuinn engage in a back-and-forth about the purpose and cost of Tarsnap, with hiAndrewQuinn reiterating it's primarily a hobby.
- placardloop bluntly states the markup: "The markup on tarsnap is wild."
- iuno points out a potential typo in a referenced cost estimator.