- It's just for the "S3 Express One Zone" bucket class, which is more expensive (16c/GB/month compared to 2.3c for S3 standard tier) and less highly available, since it lives in just one availability zone
- "With each successful append operation, you create a part of the object and each object can have up to 10,000 parts. This means you can append data to an object up to 10,000 times."
That 10,000 parts limit means this isn't quite the solution for writing log files directly to S3.
Azure supports 50,000 parts, zone-redundancy, and append blobs are supported in the normal "Hot" tier, which is their low-budget mechanical drive storage.
Note that both 10K and 50K parts means that you can use a single blob to store a day's worth of logs and flush every minute (1,440 parts). Conversely, hourly blobs can support flushing every second (3,600 parts). Neither support daily blobs with per-second flushing for a whole day (86,400 parts).
Typical designs involve a per-server log, per hour. So the blob path looks like:
This seems insane, but it's not a file system! You don't need to create directories, and you're not supposed to read these using VIM, Notepad, or whatever.
The typical workflow is to run a daily consolidation into an indexed columnstore format like Parquet, or send it off to Splunk, Log Analytics, or whatever...
Microsoft did this by sacrificing other features of object storage that S3 and GS had since the beginning, primarily performance, automatic scaling, unlimited provisioning and cross-sectional (region wide) bandwidth. Azure blob storage did not have parity on those features back in 2015 and data platform applications could not be implemented on top of it as a result. Since then they fixed some of these, but there are still areas where Azure lacks scaling features that are taken for granted on AWS and GCP.
If I need to consolidate anyway, is this really a win for this use case? I could just upload with {hour}_{minute}.txt instead of appending every minute, right?
> Wow, I'm surprised it took AWS this long to (mostly) catch up to Azure, which had this feature back in 2015:
Microsoft had the benefit of starting later and learning from Amazon's failures and successes. S3 dates from 2006.
That being said, both Microsoft and Google learned a lot, but also failed at learning different things.
GCP has a lovely global network, which makes multi-region easy. But they spent way too much time on GCE and lost the early advantage they had with Google App Engine.
Azure severely lacks in security (check out how many critical cross-tenant security vulnerabilities they've had in the past few years) and reliability (how many times have there been various outages due to a single DC in Texas failing; availability zones still aren't the default there).
AWS ranks features based on potential income from customers. Normally there’s a fairly big customer PFR needed to get a service team to implement a new feature.
I always found it strange that AWS seems to have 2-3x as many products or services as Azure, but it has these bizarre feature gaps where as an Azure user I think: "Really? Now? In this year you're finally getting this?"
(Conversely, Azure's low-level performance is woeful in comparison to AWS and they're still slow-walking the rollout of their vaguely equivalent networking and storage called Azure Boost.)
I've only used azure a little bit, and mostly liked it - but I'd love to know what kinds of things you're referring to here (mostly on AWS only, so probably I don't even know what I'm missing out on).
What Azure has that from what I've seen AWS does not:
Resource Groups that actually act like folders, not just as special tags.
Resources with human-readable names instead of gibberish identifiers.
Cross-region and cross-subscription (equiv. to AWS account) views of all resources as the default, not as a special feature.
Single pane-of-glass across all products instead of separate URLs and consoles for each thing. E.g.: a VM writing to an S3 bucket dedicated to it are "far apart" from each other in AWS consoles, but the equivalent resources are directly adjacent to each other in an Azure Resource Group when viewed in its Portal.
Azure Application Insights is a genuinely good APM, and the Log Analytics workspace it uses under the hood is the consistent log collection platform across everything in Azure and even Entra ID and parts of Microsoft 365. It's not as featureful as Splunk, but the query language is up there in capability.
Azure App Service has its flaws, but it's by far the most featureful serverless web app hosting platform.
Don’t forget, you don’t pay for a stopped vm in azure! You only pay while it is running. This makes things like dev environments much more affordable, since you won’t be paying for nights/weekends.
The original title is "Amazon S3 Express One Zone now supports the ability to append data to an object" and the difference is extremely important! I was excited for a moment.
For comparison, while GCS doesn't support appends directly, there's hacky but effective workaround in that you can compose existing objects together into new objects, without having to read & write the data. If you have existing object A, upload new object B, and compose A and B together so that the resulting object is also called A, this effectively functions the same as appending B into A.
Colossus allows appends, so it would make sense that there's a backdoor way to take advantage of that in GCS. It seemed silly of me that Google didn't just directly allow appends given the architecture.
There are some limitations[0] to work around (can only compose 32 at a time and it doesn't auto delete composed parts), but I find this approach super useful for data ingest and ETL processing flows while being quite easy to use.
It's similar but no really the same thing. It has to be done up front by initiating a multi-part upload to start. The parts are still technically accessible as S3 objects but through a different API. But the biggest limitation is that each part has to be >5MB (except for the final part)
It's crazy to me that anyone would still consider S3 after R2 was made available, given the egress fees. I regularly see people switching to R2 and saving thousands or hundreds of thousands by switching.
> It's crazy to me that anyone would still consider S3 after R2 was made available, given the egress fees.
If your compute is on AWS, using R2 (or anything outside of AWS) for object storage means you pay AWS egress for “in-system” operations rather than at the system boundary, which is often much more expensive (plus, you also probably add a bunch of latency compared to staying on AWS infra.) And unless you are exposing your object store directly externally as your interface to the world, you still pay AWS egress at the boundary.
Now, if all you use AWS for is S3, R2 may be, from a cost perspective, a no brainer, but who does that?
For the most part I agree, but we have found that R2 does not handle large files (hundreds of GB or larger) very well. It will often silently fail with nothing being returned, so it’s not possible to handle it gracefully.
Depends a bit on your use case. If you've got lots of other infra on AWS and you don't need to store that much data then the downside of using another platform can outweigh the cost savings.
doesn't s3 have 'free' (subsidized!) transfer to other products like ec2 though? it might look better to businesspeople that "well, we're doing this processing on ec2, why not keep our objects in s3 since it's a bit cheaper!"
In most cases S3 data is not directly exposed to the client. If the middleware is EC2, then you need to pay the same egress fee, but you you will have lower latency with S3, as EC2 shares the same datacenter as S3.
This doesnt seem very useful for many cases, given that you NEED to specify a write offset in order for it to work. So you need to either track the size (which becomes more complex if you have multiple writers), or need to first request the size every time you want to do a write and then race for it using the checksum of the current object... Urghhh.
I'm curious on the different use cases for this? Firehose/kinesis whatever the name seems to have the append case covered in ways that I would think has fewer foot guns?
I know the whole point of cloud services is to pick and choose, but in general I wouldn’t express “outrage” or scoff when comparing Azure to AWS. I recommend Azure to the smallest and leanest of shops, but when you compare functionality matrices and maturity Azure is a children’s toy.
To compare the other way, Azure write blocks target replication blob containers. I consider that a primitive and yet they just outright say you can’t do it. When I engaged our TPM on this we were just told our expectations were wrong and we were thinking about the problem wrong.
The goal of my question was about what are the differences between the two solutions: I know HN is a place where I can read technical arguments based on actual experience.
Will be exciting to see what adaptations are needed and how performance and cost changes for delta lake and iceberg and other cloud mutable data storage formats. It could be really dramatic!
S3 is often used as a lowest common denominator, and a lot of the features of azure and gcs aren’t leveraged by libraries and formats that try to be cross platform so only want to expose features that are available everywhere.
If these days all object stores do append then perhaps all the data storage formats and libs can start leveraging it?
Ingress pricing is indeed cheaper. POST is $0.005 per thousand requests on standard and $0.0025 on express one.
Egress and storage however are more expensive on express one than any other tier. For comparison, glacier (instant), standard and express are $0.004, $0.023 and $0.16 per GB. Although slight, standard tier also receives additional discounts above 50 TB.
This is a fantastic addition to Amazon S3 Express One Zone! The ability to directly append data to existing objects opens up new possibilities for real-time data processing applications. Whether it's continuously adding log entries or appending video segments in a media workflow, this feature will streamline workflows and improve efficiency for many use cases. It's great to see AWS continuing to innovate and make data management even more flexible and user-friendly. Excited to see how this feature enhances the scalability of applications across various industries!
I don't understand the bashing in the comments. I image this is a tough distributed systems challenge (as with anything S3). Of course AWS is charging more since they've cracked it.
Does anybody know if appending still has that 5TB file limit ?
Livestream is not usually done by writing to/reeding from a single media file. Instead, the media is broken into few second long segments. The current set of segments are indicated by a HLS and/or DASH manifest, which is updated as new segments appear.
But most of them are compliant with the standard S3 API. If I use AWS SDK to write data to my on-prem Ceph/Minio/SeaweedFS/Hitachi storage, I want this SDK to support the concept of appending data to an object.
Incredible breakthrough. What will they come up with next the ability to remove data from an object? It’s clear that not working from home is really working out for them
Amazon has no right to do this - it no longer owns the S3 standard and should respect the ecosystem and community.
S3 has stagnated for a long time, allowing it to become a standard.
Third parties have cloned the storage service and a vast array of software is compatible. There’s drivers, there’s file transfer programs and utilities.
What does it mean that Amazon is now changing it.
Does Amazon even really own the standard any more, does it have the right to break the long standing standards?
I’m reminded of IBM when they broke compatibility of the PS/2 computers just so it could maintain dominance.
Wrote some notes on this here: https://simonwillison.net/2024/Nov/22/amazon-s3-append-data/
Key points:
- It's just for the "S3 Express One Zone" bucket class, which is more expensive (16c/GB/month compared to 2.3c for S3 standard tier) and less highly available, since it lives in just one availability zone
- "With each successful append operation, you create a part of the object and each object can have up to 10,000 parts. This means you can append data to an object up to 10,000 times."
That 10,000 parts limit means this isn't quite the solution for writing log files directly to S3.
Wow, I'm surprised it took AWS this long to (mostly) catch up to Azure, which had this feature back in 2015: https://learn.microsoft.com/en-us/rest/api/storageservices/u...
Azure supports 50,000 parts, zone-redundancy, and append blobs are supported in the normal "Hot" tier, which is their low-budget mechanical drive storage.
Note that both 10K and 50K parts means that you can use a single blob to store a day's worth of logs and flush every minute (1,440 parts). Conversely, hourly blobs can support flushing every second (3,600 parts). Neither support daily blobs with per-second flushing for a whole day (86,400 parts).
Typical designs involve a per-server log, per hour. So the blob path looks like:
This seems insane, but it's not a file system! You don't need to create directories, and you're not supposed to read these using VIM, Notepad, or whatever.The typical workflow is to run a daily consolidation into an indexed columnstore format like Parquet, or send it off to Splunk, Log Analytics, or whatever...
Microsoft did this by sacrificing other features of object storage that S3 and GS had since the beginning, primarily performance, automatic scaling, unlimited provisioning and cross-sectional (region wide) bandwidth. Azure blob storage did not have parity on those features back in 2015 and data platform applications could not be implemented on top of it as a result. Since then they fixed some of these, but there are still areas where Azure lacks scaling features that are taken for granted on AWS and GCP.
If I need to consolidate anyway, is this really a win for this use case? I could just upload with {hour}_{minute}.txt instead of appending every minute, right?
Consolidation is for archival cost efficiency and long-term analytics. If you don't append regularly, you can lose up to 59 minutes of data.
In all fairness. Shipping unreliable features for unreliable services is a lot easier
> Wow, I'm surprised it took AWS this long to (mostly) catch up to Azure, which had this feature back in 2015:
Microsoft had the benefit of starting later and learning from Amazon's failures and successes. S3 dates from 2006.
That being said, both Microsoft and Google learned a lot, but also failed at learning different things.
GCP has a lovely global network, which makes multi-region easy. But they spent way too much time on GCE and lost the early advantage they had with Google App Engine.
Azure severely lacks in security (check out how many critical cross-tenant security vulnerabilities they've had in the past few years) and reliability (how many times have there been various outages due to a single DC in Texas failing; availability zones still aren't the default there).
AWS ranks features based on potential income from customers. Normally there’s a fairly big customer PFR needed to get a service team to implement a new feature.
I always found it strange that AWS seems to have 2-3x as many products or services as Azure, but it has these bizarre feature gaps where as an Azure user I think: "Really? Now? In this year you're finally getting this?"
(Conversely, Azure's low-level performance is woeful in comparison to AWS and they're still slow-walking the rollout of their vaguely equivalent networking and storage called Azure Boost.)
Low level performance is underselling the issues. Blob storage is not infinitely scalable. That means it’s just not the same thing as s3.
I've only used azure a little bit, and mostly liked it - but I'd love to know what kinds of things you're referring to here (mostly on AWS only, so probably I don't even know what I'm missing out on).
What Azure has that from what I've seen AWS does not:
Resource Groups that actually act like folders, not just as special tags.
Resources with human-readable names instead of gibberish identifiers.
Cross-region and cross-subscription (equiv. to AWS account) views of all resources as the default, not as a special feature.
Single pane-of-glass across all products instead of separate URLs and consoles for each thing. E.g.: a VM writing to an S3 bucket dedicated to it are "far apart" from each other in AWS consoles, but the equivalent resources are directly adjacent to each other in an Azure Resource Group when viewed in its Portal.
Azure Application Insights is a genuinely good APM, and the Log Analytics workspace it uses under the hood is the consistent log collection platform across everything in Azure and even Entra ID and parts of Microsoft 365. It's not as featureful as Splunk, but the query language is up there in capability.
Azure App Service has its flaws, but it's by far the most featureful serverless web app hosting platform.
Etc...
Don’t forget, you don’t pay for a stopped vm in azure! You only pay while it is running. This makes things like dev environments much more affordable, since you won’t be paying for nights/weekends.
Kusto is wonderful. I'd love to be able to use outside of log analytics.
Not directly, but enough to write once every hour for more than a year!
Yeah, or I guess log rotation will work well - you can write 10,000 lines to one key and then switch to a new key name.
This will require some serious buffering.
The original title is "Amazon S3 Express One Zone now supports the ability to append data to an object" and the difference is extremely important! I was excited for a moment.
For comparison, while GCS doesn't support appends directly, there's hacky but effective workaround in that you can compose existing objects together into new objects, without having to read & write the data. If you have existing object A, upload new object B, and compose A and B together so that the resulting object is also called A, this effectively functions the same as appending B into A.
https://cloud.google.com/storage/docs/composite-objects#appe...
Colossus allows appends, so it would make sense that there's a backdoor way to take advantage of that in GCS. It seemed silly of me that Google didn't just directly allow appends given the architecture.
There are some limitations[0] to work around (can only compose 32 at a time and it doesn't auto delete composed parts), but I find this approach super useful for data ingest and ETL processing flows while being quite easy to use.
[0] https://chrlschn.dev/blog/2024/07/merging-objects-in-google-...
S3 can also do this https://docs.aws.amazon.com/AmazonS3/latest/API/API_UploadPa...
It's similar but no really the same thing. It has to be done up front by initiating a multi-part upload to start. The parts are still technically accessible as S3 objects but through a different API. But the biggest limitation is that each part has to be >5MB (except for the final part)
It's totally different thing and requires special way to initiate multi-part uploading.
totally different how?
It's crazy to me that anyone would still consider S3 after R2 was made available, given the egress fees. I regularly see people switching to R2 and saving thousands or hundreds of thousands by switching.
> It's crazy to me that anyone would still consider S3 after R2 was made available, given the egress fees.
If your compute is on AWS, using R2 (or anything outside of AWS) for object storage means you pay AWS egress for “in-system” operations rather than at the system boundary, which is often much more expensive (plus, you also probably add a bunch of latency compared to staying on AWS infra.) And unless you are exposing your object store directly externally as your interface to the world, you still pay AWS egress at the boundary.
Now, if all you use AWS for is S3, R2 may be, from a cost perspective, a no brainer, but who does that?
For the most part I agree, but we have found that R2 does not handle large files (hundreds of GB or larger) very well. It will often silently fail with nothing being returned, so it’s not possible to handle it gracefully.
Depends a bit on your use case. If you've got lots of other infra on AWS and you don't need to store that much data then the downside of using another platform can outweigh the cost savings.
doesn't s3 have 'free' (subsidized!) transfer to other products like ec2 though? it might look better to businesspeople that "well, we're doing this processing on ec2, why not keep our objects in s3 since it's a bit cheaper!"
S3 has free data transfer within the same region.
People still use S3 because doing business with Cloudflare is a liability.
Genuinely curious what you meant by this?
Elaborate?
There are reasons like this: https://github.com/cloudflare/terraform-provider-cloudflare/...
In most cases S3 data is not directly exposed to the client. If the middleware is EC2, then you need to pay the same egress fee, but you you will have lower latency with S3, as EC2 shares the same datacenter as S3.
Just wait until Cloudflare decides you can afford to be on a different plan.
This doesnt seem very useful for many cases, given that you NEED to specify a write offset in order for it to work. So you need to either track the size (which becomes more complex if you have multiple writers), or need to first request the size every time you want to do a write and then race for it using the checksum of the current object... Urghhh.
Please fix title: Amazon S3 Express One Zone now supports the ability to append data to an object
I'm curious on the different use cases for this? Firehose/kinesis whatever the name seems to have the append case covered in ways that I would think has fewer foot guns?
I am surprised it was not supported until now? How does it compare to azure blob append (which exists for years)?
I have been using azure storage append blob to store logs of long running tasks with periodic flush (see https://learn.microsoft.com/en-us/rest/api/storageservices/u...)
I know the whole point of cloud services is to pick and choose, but in general I wouldn’t express “outrage” or scoff when comparing Azure to AWS. I recommend Azure to the smallest and leanest of shops, but when you compare functionality matrices and maturity Azure is a children’s toy.
To compare the other way, Azure write blocks target replication blob containers. I consider that a primitive and yet they just outright say you can’t do it. When I engaged our TPM on this we were just told our expectations were wrong and we were thinking about the problem wrong.
I did not want to express any outrage (even sarcastically), just surprise and the fact that I don't know very well the AWS offer.
> Azure write blocks target replication blob containers
I am sorry but what does it mean?
The goal of my question was about what are the differences between the two solutions: I know HN is a place where I can read technical arguments based on actual experience.
This is specifically for S3 "Express One Zone"
Logs are a terrible usecase for this. Loki already existed and it uses the cheaper more highly available s3
I wonder at which point they'll admit they've added back all the complexity of a hierarchical filesystem.
Will be exciting to see what adaptations are needed and how performance and cost changes for delta lake and iceberg and other cloud mutable data storage formats. It could be really dramatic!
S3 is often used as a lowest common denominator, and a lot of the features of azure and gcs aren’t leveraged by libraries and formats that try to be cross platform so only want to expose features that are available everywhere.
If these days all object stores do append then perhaps all the data storage formats and libs can start leveraging it?
I got excited until I saw the one zone part. That is a critical difference in terms of cost.
Isn't it cheaper than normal S3 though?
> S3 Express One Zone delivers data access speed up to 10x faster and request costs up to 50% lower than S3 Standard [0]
The critical difference seems to be in availability (1 AZ)
[0] https://aws.amazon.com/s3/storage-classes/express-one-zone/
Storage cost is much higher. Express One Zone is $0.16 per GB while standard S3 is $0.023 per GB ( https://aws.amazon.com/s3/pricing/?nc=sn&loc=4 )
Ingress pricing is indeed cheaper. POST is $0.005 per thousand requests on standard and $0.0025 on express one.
Egress and storage however are more expensive on express one than any other tier. For comparison, glacier (instant), standard and express are $0.004, $0.023 and $0.16 per GB. Although slight, standard tier also receives additional discounts above 50 TB.
> Compared to S3 Standard, even though the new class offers 50% cheaper request pricing—storage is almost seven times more expensive.
The house always wins https://www.vantage.sh/blog/amazon-s3-express-one-zone
I wonder what the implications for all the s3-like APIs is going to be.
Good point - plus the conditional writes recently announced. Did anyone else implement that?
MinIO actually supported conditional writes a year before S3
This is a fantastic addition to Amazon S3 Express One Zone! The ability to directly append data to existing objects opens up new possibilities for real-time data processing applications. Whether it's continuously adding log entries or appending video segments in a media workflow, this feature will streamline workflows and improve efficiency for many use cases. It's great to see AWS continuing to innovate and make data management even more flexible and user-friendly. Excited to see how this feature enhances the scalability of applications across various industries!
This sounds like a thin wrapper over an underlying object store
I don't understand the bashing in the comments. I image this is a tough distributed systems challenge (as with anything S3). Of course AWS is charging more since they've cracked it.
Does anybody know if appending still has that 5TB file limit ?
Does it really work for livestreams? Can I stream read and write on the same video file? That is huge if true!
Edit: oh it’s only in one AZ
Livestream is not usually done by writing to/reeding from a single media file. Instead, the media is broken into few second long segments. The current set of segments are indicated by a HLS and/or DASH manifest, which is updated as new segments appear.
There are many, many S3 compatible storage services out there provided by other companies.
Most of them cheaper, some MUCH cheaper.
But most of them are compliant with the standard S3 API. If I use AWS SDK to write data to my on-prem Ceph/Minio/SeaweedFS/Hitachi storage, I want this SDK to support the concept of appending data to an object.
If you want to know the differences between Express One Zone and normal, check https://www.vantage.sh/blog/amazon-s3-express-one-zone this blog post. I had no idea this even existed. tl;dr: it's x7 expensive.
Incredible breakthrough. What will they come up with next the ability to remove data from an object? It’s clear that not working from home is really working out for them
Amazon has no right to do this - it no longer owns the S3 standard and should respect the ecosystem and community.
S3 has stagnated for a long time, allowing it to become a standard.
Third parties have cloned the storage service and a vast array of software is compatible. There’s drivers, there’s file transfer programs and utilities.
What does it mean that Amazon is now changing it.
Does Amazon even really own the standard any more, does it have the right to break the long standing standards?
I’m reminded of IBM when they broke compatibility of the PS/2 computers just so it could maintain dominance.