Deployment strategy

By: ved singh user 13 Jun 2017 at 10:35 p.m. CDT

7 Responses
ved singh gravatar
Hi, I'm currently contemplating on the deployment design for Gluu server taking in to account concerns like failover,high availability and load balancing. This is sketch which kind of depicts the planned deployment view. Note - Users are stored in a RDBMS system [https://drive.google.com/file/d/0B2f0ri6BmhOnakkzdXlfN0M0dk0/view?usp=sharing](https://drive.google.com/file/d/0B2f0ri6BmhOnakkzdXlfN0M0dk0/view) My questions are: 1) How to functionally detect a Gluu server unavailability? Process not running in one brute way.Is there an url (like https://serverhome/identity/home.htm) which return http 200 response which can be used as a health check url? 2) How or what is the best approach to ensure that the local ldap storage of Gluu is in sync with each other ? 3) Should an external LDAP be considered in the design and implement external-internal LDAP sync via cache-refresh mechanism for a more robust solution. Thanking in advance for any thought or inputs. Ved
CentOS72
closed

Answers

By ved singh user 13 Jun 2017 at 10:46 p.m. CDT

Copy
ved singh gravatar
Hi, I believe clustering strategy in 3.0.1 is not present. reference thread : https://support.gluu.org/installation/4033/csync2-clustering-support-in-gluu-301/ Or I'm missing something here? Ved

By Mohib Zico staff 14 Jun 2017 at 1:24 a.m. CDT

Copy
Mohib Zico gravatar
Hi Ved, [Here](https://gluu.org/docs/cm/) is the cluster configuration doc for Gluu Server v3.

By ved singh user 14 Jun 2017 at 7:10 a.m. CDT

Copy
ved singh gravatar
Thanks Mohib. I was not sure if clustering strategy is fully available at this time in version 3.0.1 as the documentation version says alpha. As regards to my first question, is there any url or log statement(s) which can indicate health of the Gluu server. Thanks Ved

By Mohib Zico staff 14 Jun 2017 at 10:26 a.m. CDT

Copy
Mohib Zico gravatar
>> As regards to my first question, is there any url or log statement(s) which can indicate health of the Gluu server. You can achieve such in this way: - Create a file.. something like 'monitoring.html' and put them inside /var/www/html/ of every Gluu Server. - Then start monitoring them from your load balancer.

By ved singh user 14 Jun 2017 at 2:33 p.m. CDT

Copy
ved singh gravatar
Thanks Mohib.

By Vitalii Bendyna user 22 Jun 2017 at 4:50 a.m. CDT

Copy
Vitalii Bendyna gravatar
> Create a file.. something like 'monitoring.html' and put them inside /var/www/html/ of every Gluu Server But this case will only cover the availability of Apache web server, not application itself. How I can provide monitoring of health for internal application's parts? Thanks.

By ved singh user 22 Jun 2017 at 8:18 a.m. CDT

Copy
ved singh gravatar
That exactly what my confusion was. We can check whether Gluu process is running as per the process which Mohib mentioned, but how do we ensure that the internal health is kosher. For example - A new user authentication session was created but it replication to other nodes fails because of underlying issues(can my many reason there). How do we check for other such cases. Thanks Ved

Post an answer