Ceph Deduplication (14.2.4)

[paradox55]

Can we get a guide made for deduplication on 14.2.4?

Following the examples here I was able to get a basic setup going, which wasn't possible on 14.2.2.

https://docs.ceph.com/docs/master/dev/deduplication/

Thing is, I don't really know what they want me to put in for offsets. And I believe foo and foo-chunk are custom scripts, which aren't mentioned anywhere in the guide.

But the commands all work from my testing.

 

[Alwin]

Can we get a guide made for deduplication on 14.2.4?

This doesn't seem to be published yet, no 'dedup' in the release notes and I can't find a 'ceph-dedup-tool' in Ceph 14.2.4.

 

[paradox55]

This doesn't seem to be published yet, no 'dedup' in the release notes and I can't find a 'ceph-dedup-tool' in Ceph 14.2.4.

iirc the dedup tool is just for finding the optimal offsets and fixing the reference count.

The actual commands for enabling deduplication worked for me so the base functionality is there.

 

[Alwin]

iirc the dedup tool is just for finding the optimal offsets and fixing the reference count.

Yes but isn't it a good indication that the feature released / production ready yet? And it only exists in the developer documentation.

 

[paradox55]

Yes but isn't it a good indication that the feature released / production ready yet? And it only exists in the developer documentation.

I have reviewed the source code of 14.2.4. Deduplication is supported. The ceph-dedup-tool is in 14.2.4, but has to be built with debug/testing flags.

Well, it looks supported to me anyway. Lots of stuff in there regarding deduplication - and this is a production build, not developer.

Edit: Might just be reading it wrong, too. Does VDO work with CephFS or does it have to be on the underlying block device?

 

[paradox55]

Using the instructions in the source code I was able to get dedup to work correctly, so I'm happy about that!

root@cl-01:~# rados -p dedup_base put foo ./test_lol

root@cl-01:~# rados -p dedup_base put foo ./test_lol2


root@cl-01:~# rados ls -p dedup_chunk

bar-chunk

foo-chunk

99db1422af30c1857cc0c45f8f5ea7418c397d8d


root@cl-01:~# rados ls -p dedup_base

foo

bar


root@cl-01:~# echo -n "There hi" | sha1sum

99db1422af30c1857cc0c45f8f5ea7418c397d8d  -

 

[paradox55]

Incidentally the reason the dedup tool is in testing is because it fails to build in rpm packages otherwise. Iirc it's stable otherwise.

 

[Alwin]

It still seems to me not much production ready / supported and I will still advise against using it. I hope it will work for you and you will not lose any data. If you don't mind, would you post some experience report later on? Thanks.

 

[jsterr]

Is there a way to use dedup with ceph and proxmox?

 

[jw6677]

As this is a top performing page in Google:

VDO appears to work on top of RBD volumes in my testing. I had some trouble with boot ordering, but seems fine otherwise.

If you're working on this stuff for fun, a word of caution, be very very careful to keep duplicate copies of your VDO data for a long while. When working with VDO, several times along the journey i've hit issues which resulted in losing whole volumes of data.
RHEL has documented VDO well, so setting it up is easy, but the number of folk using it outside of production environments doesn't seem to have saturated widely enough to allow for easily finding debugging help online.
--> If you run into an issue, there is a good chance you'll be on your own in fixing it, has been my experience. If anyone has a resource for debugging VDO as a hobby user, I am all ears!

 

[itNGO]

Has this made any progress towards being something supported? Is it in newer CEPH-Versions?

 

[jw6677]

Has this made any progress towards being something supported? Is it in newer CEPH-Versions?

@itNGO , Not that I have seen.

 

[itNGO]

@itNGO , Not that I have seen.

Again a year later. Anything new about this?

 

[bmernz]

From https://docs.ceph.com/en/latest/dev/deduplication/#how-to-use-deduplication :

• This feature is highly experimental and is subject to change or removal.