Frame link module not connected to standby

[cer113]

HI

I pulled both Frame Link modules in one Frame and then after that I had error not connected to standby. Went to console and try to factory reset this didnt help. Then I removed standby appliance and now there is no way to add this secondary Frame Link module to be standby.

Apparently I shouldnt remove both Frame link modules from one frame because this looks it is not recoverable

I found this

https://techlibrary.hpe.com/docs/synergy/composer2/msg/s_hacluster-action-activating-standby-appliance-atlas.html#:~:text=From%20the%20main%20menu%2C%20select,Select%20Actions%20%3E%20Activate%20standby.

But i could never find any way to activate standby not in one view not in Maintenance console

Looks like other one is just destroyed by pulling both Frame Link modules from one frame

When i go to maintenance console it says it is standby but OneView cannot see that

Is there any way to recover from this

[cer113]

do I have to reimage I basically did FLM factory reset but One view cannot see it. But in Maintenance console it shows clearly and it says it is OK

[cer113]

tried to remove standby appliance and now OneView cannot see it i just dont know where to add it back

[cer113]

It is there but for some reason I cannot activate secondary Frame Link module although everything is green and Maintenance console can see it

[cer113]

based on this as soon as I put it back it should restore but nothing that is described in these documents works

https://support.hpe.com/hpesc/public/docDisplay?docLocale=en_US&docId=a00101499en_us&page=GUID-A34CC01C-7852-4868-B65D-37137920A038.html

[cer113]

this is the one that lost connection and OneView says it is there but for some reason I cannot reestablish high availability. There is no way to activate it

[cer113]

I Factory reset standby Frame Link module and that didnt do anything.

Do I have start all from scratch factory reset everything  since this is the only thing i didnt do

[DanCernese]

I may be incorrect but I believe you are mixing terminology between Composers and Frame Link Modules.  References to high availbility of the management ring have nothing to do with high availability of the composers or whether they are standby or actice.   Having pulled two FLMs out though, things will get messy.  I do not believe this is simple or straight-forward enough to assist you in this forum.  I would either open a support case or yes, reset everything and start over (assuming you are not losing anything).

[cer113]

I could do that but where to start. I already factory reset both FLMs and nothing everything came back the same

I didnt restart appliance not sure if that would help

[cer113]

Here it says that they are linked which is most confusing

When I started testing i just pulled one FLM at a time and it was my mistake pulling both and apparently that is catastrophic faiure that cannot be recovered from either from one view or Maintenance console

[DanCernese]

---

@cer113 wrote:

I could do that but where to start. I already factory reset both FLMs and nothing everything came back the same

I didnt restart appliance not sure if that would help

---

You cannot just do one thing.  If you have factory reset both FLMs (not just reset but factory default), then you must connect to each Composer console and see if there is an error message to resolve.  After resolving that, you can connect to the active Composer and see if it has formed a cluster with the second composer.  If you have issue still-- you should factory reset those composer(s).  ...but I'm just guessing.

[DanCernese]

---

@cer113 wrote:

Here it says that they are linked which is most confusing

Why is this confusing?  It is exactly as it should be, especially after reset.  Again-- do not confuse the FLMs with the Composers' status.

[cer113]

This scenario was not covered in documentation. After redundancy is lost there is no way to recover it your FLMS are lost then they are lost forever

OK I am just testing this but I wouldnt feel putting in production and have all these issues

I guess we will try Cisco UCS see how is this one responding  or  IBM

[DanCernese]

It is not appropriate to assume from your actions that anything is lost forever.  You have confused terminology and have been randomly experimenting without understanding the fundamental building blocks.  There is no loss of FLM redundancy now that it has recovered.  You have not shown or discussed recovering the Composer configuration.  That is why I recommended that if you truly want to make this configuration operational you learn how to reset all the components or open a support case.  All in all, you would benefit greatly from training on the purpose of the management ring and composer cluster.

[cer113]

now this is the error

SYN_02_Enclosure, appliance bay 1 cannot join the appliance cluster because it belongs to another cluster or is trying to form one. There can be only one appliance cluster managing the linked frames.

and this should fix it

https://techlibrary.hpe.com/docs/iss/shared/error_gen10/GUID-2466EB2D-7DE1-4461-8FC7-A9F081CFB7EB.html

[thutchings]

Hello,

Agree with Dan. As far as can be seen, the FLM's are working fine. When the communication from one Composer was lost to another it caused the cluster that is formed between them to lose sync. Usually, just restoring communications is enough to get them to sync up again.

Have you tried restarting the standby appliance that is not joining the cluster? You can do that by SSH'ing to the maintenance console and selecting the restart option from there. That is sometimes all that is required to kick off the synchronization. If it doesn't start to join the cluster after that then you may need to factory reset the standby composer. You can also reimage the standby, but just a factory reset of the standby should be enough to get it to join the cluster again and sync the data. I would make sure that when selecting to factory reset the standby (or reimage - if you go that route) I would make sure it is the correct Composer by confirming which one is active in the GUI and performing the actions against the other one.

Regards

[cer113]

Thanks for all your help I reimaged standby and then factory reset this initialized joining standby composer to cluster

Much appreciated

[cer113 · Accepted solution]

Just when I was about to finish this another alert came

The appliance has been started under circumstances where the database may not match the managed environment. To prevent unintentional allocation of duplicate MAC address, world wide names and serial numbers, existing ranges have been disabled. For these identifiers, replacement ranges have been created. If none of these identifiers have been allocated since the time when the database matched the managed environment, it is safe to re-enable the range and remove the replacement range. When keeping a replacement range, ensure it does not conflict with addresses or identifiers used in the environment.

I removed newly generated WWNs MAC addresses but this alert wont go away

What needs to be done to remove this alert

HP tech tells me i dont know will ask around anhyone had this case before