Skip to main content

A journey of Scaling generation of file system snapshots

We are a cloud file system startup managing billions of files for thousands of customers and one of the constant scaling issue is "how do you generate a snapshot of cloud filesystem" and send it to the client to start an initial sync. To elaborate more let say a customer has so far uploaded 25M files. Now he starts a new office and wants to setup a Netgear/NAS appliance and install our server sync software. Before the Netgear/NAS appliance could start syncing it needs a consistent snapshot of cloud filesystem so it could start that as a starting point before it could start syncing  changes based on events.  Now 25M is a big number right now but 5 year back the same problem was with if a customer has 500K files and a new employee joins and installs a sync software on his laptop, how do you send the sync client a consistent server snapshot.

It seems every 1-2 year or so we solve this problem and as scale increases we have to come up with something new to make it more faster than before.

Snapshot solution Version 1

When we initially started we were using filers and using NFS and when a customer would add a folder on local or from web ui, we on our server would replicate the operation via NFS.   When  a new user sync install would request a snapshot we would crawl the file system and send him a snapshot.  This worked for sometime but crawling the file system is slow and NFS would freeze the OS.

Snapshot solution Version 2

To solve the crawling issue we started a python process on the filer that would start with an initial image via crawling and it would hold it in memory. It would then listen to INotify events when we would do operation via NFS and it would update the  in memory snapshot.  When a new user sync install would request the snapshot we would serve it from the cached in memory install.  This worked great for sometime but it ran into issues whenever we would bring restart a filer or the python process would restart or the snasphot was so big that it would cause OOM.


Snapshot solution Version 3

We ditched NFS and we started storing filesystem metadata in a NOSQL database called as BerkelyDB. Its a key value database and we would store the file on filer but BerkelyDB would store the metadata about file XYZ belongs to folder ABC and is located on ZZZ filer at this location. When a new user sync install would request the snapshot we would just dump all BerkelyDB data(paths were stored as key and folder metadata as values) and generate  a snapshot.  This worked for a long time until we started reaching customers with 3M+ files and this would buckle under pressure if multiple clients would request snapshots at same time. BerkelyDB would run into memory issues or cache thrashing.


Snapshot solution Version4

Besides cache/memory issues  BerkelyDB's biggest issue was data loss and replication. So we ditched BerkelyDB and started using Mysql. We created Folder,File,Version table and 1000s of shards to store metadata. We are right now running 100+ mysql servers to store this data. One customer's data was stored in one shard so whenever a user sync install would request the snapshot we would join the three tables and spit out the snapshot.  This was one big breakthrough from other areas of application also.  But one big problem with this approach was that we allow customers to upload any number of versions and the snapshot only cares about latest version of the file. So in order to derive latest version of the file we had to discard old versions using a correlated subquery.  So this works fine for 5M files but it really buckled up under 8M+ files as we had to fire a correlated subquery for each file.

Snapshot solution Version5
We found out that if we joined Folder,File,version table and then use unix sort to sort the file and then use python to filter out latest version it was much faster than correlated queries. But this also buckled under pressure when we hit a customer with 25M files. The join in mysql db takes 2 hours to spit the data and the unix sort takes 2-3 hours.  Also the snapshot is large so we started caching this on disk and we used nginx as a reverse proxy to start serving cached snapshot. But again lately some big customers started facing issues with snapshot timeouts.


Snapshot solution Version6
This weekend we are again improvising on this solution as we found out that if we could denormalize the latest version when the file is being added,copied, moved,deleted,restored on the File table record, then to generate a snapshot we just need to join folder,file table and serve it.  From our tests we found out that generating a snapshot on 25M file customer takes only 6 minutes compared to 5 hour process with join query and unix sort approach.  Off course this has twice the space complexity but the time complexity is way better.

Offcourse this requires a lot of grunt work to modify the old code and we would need to migrate billions of files so all this would rollout slowly over the course of this month but I am hoping this solution would again buy us some time.

The core problem is snapshot and we need to find out a better way to bootstrap a client instead of sending him the entire snapshot so may be this saga has more evolution than what we had seen in the past :).

Comments

Popular posts from this blog

Haproxy and tomcat JSESSIONID

One of the biggest problems I have been trying to solve at our startup is to put our tomcat nodes in HA mode. Right now if a customer comes, he lands on to a node and remains there forever. This has two major issues: 1) We have to overprovision each node with ability to handle worse case capacity. 2) If two or three high profile customers lands on to same node then we need to move them manually. 3) We need to cut over new nodes and we already have over 100+ nodes.  Its a pain managing these nodes and I waste lot of my time in chasing node specific issues. I loath when I know I have to chase this env issue. I really hate human intervention as if it were up to me I would just automate thing and just enjoy the fruits of automation and spend quality time on major issues rather than mundane task,call me lazy but thats a good quality. So Finally now I am at a stage where I can put nodes behing HAProxy in QA env. today we were testing the HA config and first problem I immediat...

Adding Jitter to cache layer

Thundering herd is an issue common to webapp that rely on heavy caching where if lots of items expire at the same time due to a server restart or temporal event, then suddenly lots of calls will go to database at same time. This can even bring down the database in extreme cases. I wont go into much detail but the app need to do two things solve this issue. 1) Add consistent hashing to cache layer : This way when a memcache server is added/removed from the pool, entire cache is not invalidated.  We use memcahe from both python and Java layer and I still have to find a consistent caching solution that is portable across both languages. hash_ring and spymemcached both use different points for server so need to read/test more. 2) Add a jitter to cache or randomise the expiry time: We expire long term cache  records every 8 hours after that key was added and short term cache expiry is 2 hours. As our customers usually comes to work in morning and access the cloud file server it ...

Spring 3.2 quartz 2.1 Jobs added with no trigger must be durable.

I am trying to enable HA on nodes and in that process I found that in a two test node setup a job that has a frequency of 10 sec was running into deadlock. So I tried upgrading from Quartz 1.8 to 2.1 by following the migration guide but I ran into an exception that says "Jobs added with no trigger must be durable.". After looking into spring and Quartz code I figured out that now Quartz is more strict and earlier the scheduler.addJob had a replace parameter which if passed to true would skip the durable check, in latest quartz this is fixed but spring hasnt caught up to this. So what do you do, well I jsut inherited the factory and set durability to true and use that public class DurableJobDetailFactoryBean extends JobDetailFactoryBean {     public DurableJobDetailFactoryBean() {         setDurability(true);     } } and used this instead of JobDetailFactoryBean in the spring bean definition     <bean i...