Comments (4)
  • Add a Comment
  • Edit
  • More Actions v
  • Quarantine this Entry

1 1023 commented Permalink

Question. XDRP Sync configuration. <div>&nbsp;</div> After xiv volume initial copy, box DR is moving over time to 24. <br /> After the master and slave active, master and slave volume used size variation Happened. <br /> Why do these cases occur? <div>&nbsp;</div> ex) after re-sync <br /> Master site volume used size "TEST=128 GB" <br /> Slave site volume used size "DR_TEST=137 GB" <br />

2 anthonyv commented Permalink

Hi, <div>&nbsp;</div> Its a great question, but this variation is inevitable due to the fact that 'used' is a combination of real data and garbage data that has been deleted at the file system layer (at the host layer) but not at the disk layer. When we write to the master and then delete the file, the data is still on the disk and is mirrored. <br /> The used variation is not a bug and it is not a problem. The host file system will only address 'real' data which is always a subset of 'used' space.

3 1023 commented Permalink

Thanks.. <div>&nbsp;</div> But the problem, as did the raw device file system is almost certain. <br /> Sync with the concept, master and slave, does not require the same amount going to use? <br /> Another problem is, pool size is seen, the space is actually written in different master and slave <br /> Generated. What about 90GB ~ 100GB of why these things occur? How can not Reclaim ? <div>&nbsp;</div> What do customers have to explain what? <br />

4 anthonyv commented Permalink

Hi. <div>&nbsp;</div> I am struggling a little with your written English, so forgive me if my answer is not the correct response to your question. <div>&nbsp;</div> What your seeing is quite normal and is not a bug or an issue. <br /> It occurs because the local and remote machines allocate space independently of each other, so seeing variations in the 'used' sections of both the volumes and the pools is quite normal. <div>&nbsp;</div> To adequately explain how it occurs I would need to use a white board. <br /> I suggest that if your genuinely concerned, that you contact your local IBM Technical Advisor.