<html><head>
<meta name="Generator" content="Novell Groupwise Client (Version 14.2.1 Build: 124595)">
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"></head>
<body style="font: 10pt/normal Segoe UI; font-size-adjust: none; font-stretch: normal;"><div class="GroupWiseMessageBody" id="GroupWiseSection_1485242661000_Martin.Weiss@suse.com_A842D18121330000A9A50002A5DB3086_"><div>How can we verify if the policy.cfg is matching the current deployment without changing anything?</div><div>Is there some sort of a "dry-run"?</div><div><br></div><div>Regarding DeepSea vs. ceph-deploy - if we are on par or even have more functionality in DeepSea I believe the only thing that we need to add to DeepSea is a "migration method" for existing ceph-deploy based clusters before we can move away from ceph-deploy.</div><div>So getting the inventory, building the policy.cfg and then comparing DeepSea with the existing deployment with a "dry-run" would be nice to have "automated" or in a script...</div><div>Btw. I do not need a "command line editor for ceph.conf"..</div><div><br></div><div>Oh - and we need proper integration of DeepSea into existing Salt deployments i.e. in case the customer has SUSE Manager - DeepSea needs to integrate, there..</div><div><br>Thanks</div><div>Martin<br></div><div class="GroupWiseMessageBody" id="GroupWiseSection_1485242659000_Martin.Weiss@suse.com_A842D18121330000A9A50002A5DB3086_"><span class="GroupwiseReplyHeader"><br><br><span style="display: none;"></span><br></span><div><div>On Monday, January 23, 2017 12:25:01 AM Martin Weiss wrote:</div><div>> > On Thursday, January 19, 2017 07:40:28 AM Martin Weiss wrote:</div><div>> </div><div>> I have heard that we want to deprecate ceph-deploy in one of the next</div><div>> releases and replace it with deepsea. With that in mind I believe we have</div><div>> to deliver a migration path from "ceph-deploy" to "deepsea"..</div><div>> </div><div>> Ad yes - I agree - at the moment where deepsea is not yet on paar with</div><div>> ceph-deploy this might not be worth the effort... But we need to start</div><div>> building the proper solution before deprecating ceph-deploy ;-).</div><div><br></div><div>To my knowledge, the only two items ceph-deploy does that DeepSea does not is </div><div>support dm-crypt and provide a command line editor of ceph.conf. The former </div><div>is in the todo list and I am still trying to understand the requirements of </div><div>the latter.</div><div><br></div><div>> </div><div>> Oh - and btw. - this would also be an added value for customers migrating</div><div>> from NON SLES bases Ceph deployments to SES..</div><div>> > What happens if you try and something goes wrong? I don't know. My</div><div>> > </div><div>> > personal paranoia level would be high enough that I would skip using any</div><div>> > of</div><div>> > the stage orchestration commands (i.e. salt‑run state.orch ceph.stage.3)</div><div>> > and run the individual steps on the individual nodes. It would be a bit</div><div>> > tedious</div><div>> > </div><div>> > for a migration, but much easier to recover.</div><div>> </div><div>> Agreed - this adds to my wish of an automatically generated policy.cfg for</div><div>> an existing cluster.</div><div>> > In the longer term, we do have an existing card in</div><div>> > </div>> > <a href="https://github.com/SUSE/DeepSea/projects/2,">https://github.com/SUSE/DeepSea/projects/2,</a> but have not pursued it yet. <div><br></div><div>> > I</div><div>> > think the above would need to be automated for creating a policy.cfg with</div><div>> > accurate hardware profiles and somehow verifiable. Also, I do not know if</div><div>> > this</div><div>> > could be a generic solution for any Ceph cluster.</div><div>> </div><div>> So for the moment I understand this status:</div><div>> </div><div>> 1. Recommendation: Do not use DeepSea for existing ceph-deploy based</div><div>> clusters. 2. Inventory can be done with DeepSea</div><div>> 3. policy.cfg can be created manually</div><div>> 4. ?</div><div>> </div><div>> -> In case I build the policy.cfg manually and correct - what would happen</div><div>> if I go through the next steps of DeepSea, then? Will this kill / overwrite</div><div>> anything already existing?</div><div>> </div><div>If you did build your policy.cfg identically to the existing environment along </div><div>with centralizing any customized configurations (e.g. ceph.conf, rgw.conf, </div><div>etc.), then nothing would happen. Running Stage 3 and Stage 4 would check </div><div>that all roles and services are as they should be.</div><div><br></div><div>> Thanks,</div><div>> Martin</div><div>> </div><div>> > Eric</div><div>> </div><div>> _______________________________________________</div><div>> Deepsea-users mailing list</div><div>> Deepsea-users@lists.suse.com</div>> <a href="http://lists.suse.com/mailman/listinfo/deepsea-users">http://lists.suse.com/mailman/listinfo/deepsea-users</a><div><br></div></div></div></div></body></html>