q replnode DELETED filespaces

[PastaFazole]

I have these filespaces marked "DELETED" when I run 'q replnode'. The following technote suggests that these will remain until "all the chunks linking to the chunks for those deleted filespaces are gone".

http://www-01.ibm.com/support/docview.wss?uid=swg21962184

My question is: Is the deletion of these 'linking chunks' something that will happen over time, or do I/can I clear them out? I have already run an 'audit volume fix=yes'.

Thanks,
Rob



tsm: TSM1>q replnode SERVER0

Node Name Type Filespace Name FSID Files on Replication Files on
Server Server (1) Server (1)
--------------- ---- --------------- ---- ---------- --------------- ----------
SERVER0 Bkup / 6 7,297
SERVER0 Bkup /backup 7 1,501 OS1 18
SERVER0 Bkup /db 8 6 OS1 6
SERVER0 Bkup /db/arch00 10 9,411
SERVER0 Bkup /oracle 9 170,304
SERVER0 Bkup DELETED 5 0
SERVER0 Bkup DELETED 3 0

tsm: TSM1>show dedupd
Session established with server TSM1: Linux/x86_64
Server Version 7, Release 1, Level 4.102
Server date/time: 03/14/2016 09:07:03 Last access: 03/14/2016 08:00:32

**** Dedup Deletion General Status ****
Number of worker threads : 8
Number of active worker threads : 0
Number of suspended workers : 0
Number of workers to suspend : 0
Number of chunks waiting in queue : 0
Total number of chunks processed : 0

**** Dedup Deletion Worker Info ****
Worker thread 1 is not active
Worker thread 2 is not active
Worker thread 3 is not active
Worker thread 4 is not active
Worker thread 5 is not active
Worker thread 6 is not active
Worker thread 7 is not active
Worker thread 8 is not active
------------------------------------------
Total worker chunks queued : 0
Total worker chunks deleted: 0

**** Fragment Deletion Status ****
Number of busy threads: 0
Number of idle threads: 6
Number of entries queued: 0

tsm: TSM1>

 

[marclant]

My question is: Is the deletion of these 'linking chunks' something that will happen over time,

Yes, but not in every case, more on that in a minute.

or do I can I clear them out?

You cannot and should not clear them out.

I have already run an 'audit volume fix=yes'.

There is nothing broken, therefore nothing to fix.


When you backup a file using dedup, it breaks it down in chunks. So let's if NodeA backus up FileA, it will these chunks:
chunkA
chunkB
chunkC
chunkD

Now nodeB could backup also a file and this file shares these same chunks in common with FileA from NodeA. So what happens, is that for NodeB, these chunks are not stored again, but referenced to those that already exists.

Now, if you decommision nodeA, it will delete almost all references to NodeA, except for the chunks that are also referenced by other nodes. Because NodeA initially created those chunks, it will remain the owner. Now, if at some point in the future, all the chunks intially backed up by NodeA no longer are referenced by files from other nodes, meaning they expired for the other nodes, then these chunks will be dereferenced and deleted.

 

[PastaFazole]

Thanks Marclant!