| Author |
Message
|
| madrox |
 Posted: Wed Sep 05, 2018 5:39 pm Post subject: Error starting IIIB Node while Configuring IIB v10 for HA |
 |
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
I am trying to configure IIB v10 on linux for HA (dual active) and was able to set up the qmgr on both servers one following the kc
https://www.ibm.com/support/knowledgecenter/en/SSFKSJ_9.0.0/com.ibm.mq.con.doc/q018300_.html
and on server 2 after iib install ran the command for instance of broker
| Code: |
| mqsiaddbrokerinstance iibnode -e /mqha/iib |
I started the QM on server 1 with command and i see it running on both servers
but when i try to start the broker i get a message saying the
| Quote: |
BIP8059E: Command has failed because the component's Queue Manager is not available.
The command has failed because the Queue Manager associated with the component is currently unavailable.
Reissue the command after the queue manager has been restarted.
|
please advice |
|
| Back to top |
|
 |
| fjb_saper |
| Posted: Wed Sep 05, 2018 11:48 pm Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20729
Location: LI,NY
|
did you add the broker's service id to the local mqm group? on both servers?
did you add the mqm user to the mqbrkrs group on both servers? 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| exerk |
| Posted: Thu Sep 06, 2018 12:31 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
So you started QM1 on Server 1, where it will be running as the active instance, started QM1 on Server 2, where it will be running as the standby instance, then tried to connect a Broker on Server 2 to it?
I ask for clarification because that's what's implied in your statement: "...I started the QM on server 1...on server 2 after iib install ran the command for instance of broker..."
Thank you.
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 3:27 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| Quote: |
did you add the broker's service id to the local mqm group? on both servers? |
I don't know how could you elaborate. I have not done this on linux before.
| Quote: |
| did you add the mqm user to the mqbrkrs group on both servers? |
I see mqm on the mqbrkrs group on both servers when i run the command
| Code: |
| getent group mqbrkrs |
|
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 3:30 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| exerk wrote: |
So you started QM1 on Server 1, where it will be running as the active instance, started QM1 on Server 2, where it will be running as the standby instance, then tried to connect a Broker on Server 2 to it?
I ask for clarification because that's what's implied in your statement: "...I started the QM on server 1...on server 2 after iib install ran the command for instance of broker..."
Thank you. |
Sorry if i confused you. So i started the QM on server 1 and 2 and they are running.
When i try to start the broker on server 1 first and server 2 second i get the same message
| Quote: |
BIP8059E: Command has failed because the component's Queue Manager is not available.
The command has failed because the Queue Manager associated with the component is currently unavailable.
Reissue the command after the queue manager has been restarted. |
|
|
| Back to top |
|
|
| exerk |
| Posted: Thu Sep 06, 2018 3:38 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
And you followed every instruction in this KNOWLEDGE CENTRE article?
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 3:39 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| fjb_saper wrote: |
did you add the broker's service id to the local mqm group? on both servers?
did you add the mqm user to the mqbrkrs group on both servers? |
I saw your post on another thread how do you do this?
http://www.mqseries.net/phpBB2/viewtopic.php?t=74625&sid=3291d3372a140ee0d40afa79ee75dc45
| Quote: |
| The service id needs to be part of the local mqm and mqbrokers groups on both nodes!... and for a multi-instance qmgr / broker you should use domain accounts and not the SYSTEM account that is local and has a different SID on each box! ... |
|
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 4:07 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| exerk wrote: |
| And you followed every instruction in this KNOWLEDGE CENTRE article? |
Yes, it is at step # 7 where the broker is not starting.
However i have a question on step #2
| Quote: |
| On Linux and UNIX systems, create /HA/mqsi on the shared drive. Ensure that /HA/mqsi is owned by the mqbrkrs user and group, and has the access permissions rwx. The UID of the integration node user ID in /etc/passwd and the GID for mqbrkrs in /etc/group must be the same on each server. |
I have no clue on how to verify this  |
|
| Back to top |
|
|
| exerk |
| Posted: Thu Sep 06, 2018 4:53 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
From what you have posted it seems to me that you have little experience with the platform on which you are using, e.g. not knowing the basic OS commands to interrogate the system for file system ownership/permissions, or Group ID allocation. This is not a criticism of your ability, but strangers on the internet just posting the commands is not ideal - you have no way of verifying the validity of what is posted.
I suggest you engage a colleague that has that knowledge. That is a better option than relying on Google, as that colleague will be able to coach you.
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 5:45 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| exerk wrote: |
From what you have posted it seems to me that you have little experience with the platform on which you are using, e.g. not knowing the basic OS commands to interrogate the system for file system ownership/permissions, or Group ID allocation. This is not a criticism of your ability, but strangers on the internet just posting the commands is not ideal - you have no way of verifying the validity of what is posted.
I suggest you engage a colleague that has that knowledge. That is a better option than relying on Google, as that colleague will be able to coach you. |
Thank you for your suggestion! |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Thu Sep 06, 2018 7:42 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20729
Location: LI,NY
|
| Quote: |
On Linux and UNIX systems, create /HA/mqsi on the shared drive. Ensure that /HA/mqsi is owned by the mqbrkrs user and group, and has the access permissions rwx. The UID of the integration node user ID in /etc/passwd and the GID for mqbrkrs in /etc/group must be the same on each server.
If you are using an NFS v4 file server, add the following line to the /etc/exports file:
/HA * rw,sync,no_wdelay,fsid=)copy to clipboard
Start the NFS daemon by using the following command:
/etc/init.d/nfs start |
You only posted the part about the file system. However the part about userid and group id is just as important. If it is not done right you will get permission errors.
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| madrox |
| Posted: Thu Sep 06, 2018 8:34 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
Thank you all for you inputs. Turned out that i needed to create the mqha folders with the necessary accesses. Thanks again! 
..but now i realized that i actually wanted to set this up with dual active and shared file system 
anyone have any links for me to follow? |
|
| Back to top |
|
|
| exerk |
| Posted: Thu Sep 06, 2018 11:56 pm Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
| madrox wrote: |
| ..but now i realized that i actually wanted to set this up with dual active and shared file system ... |
Do you mean 'traditional' HA - single queue manager/IIB node with mountable file systems?
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| madrox |
| Posted: Fri Sep 07, 2018 5:12 am Post subject: |
|
|
Acolyte
Joined: 11 Mar 2015
Posts: 71
|
| exerk wrote: |
| madrox wrote: |
| ..but now i realized that i actually wanted to set this up with dual active and shared file system ... |
Do you mean 'traditional' HA - single queue manager/IIB node with mountable file systems? |
No i mean 2 queue managers 2 brokers, with F5. So i am assuming Queue managers will have to be clustered. |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Sep 07, 2018 5:51 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| madrox wrote: |
| No i mean 2 queue managers 2 brokers, with F5. So i am assuming Queue managers will have to be clustered. |
That's a false assumption. If you have 2 brokers behind an F5 (so I assume both brokers are hosting the same endpoints and you're load balancing) then from a broker standpoint they have separate queue managers and no connection to each other.
Now if there's something more in your architecture, for example a need for either broker to service a reply to a query then you need to start thinking about some kind of shared storage of the query status. This still doesn't have anything to do with the broker software and doesn't immediately call for a queue manager cluster.
So can you just run 2 distinct brokers behind an F5? What exactly do you have going on?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
|
|
