Feedback on CSA/OO/HCM out of the box content
R&D is planning the next releases. As part of this effort, we would like to consider your feedback on the out of the box CSA/OO/HCM content released in 2017. Can you help us with your inputs?
- What do you like the most in content?
- What is missing or can be improved in content or content tools (eg. REST Wizard)?
this is my feeling after some work to test the migration capability of a customer to the new version, I am working in a new installed from scratch 2017.08.
1 - I think this version is a good improvement for the suite, it will be easy to migrate to a new version and to scale the environment in case of needs. The concept is quite clear and by the fact that everyone is speaking and working using container I have found this solutions very interesting. I premise that I am testing the solution in one unique server, before testing it in a HA environment. There are no doubts that this is the way to proceed and to improve for the future.
2 - There are for sure some improvements that could facilitate some operativity and some that are required for a possible migration.
a) It would be nice to change the replica values of a deployments directly from the pods line entry in the ITOM console, avoiding to enter the yaml entire file everytime. It is not a big issue but change the replicas is probably the main activity in the yaml once the environment is correctly set, to change other values it is easy to enter it but for the replicas a simple field outside could be useful.
b) Encrease the size of the pods log, it happens that in case of errors you enter the log from the console and you do not have a lot of "cache" for the log, so if you go backward you do not see the entire log.
c) Maintain the number of rows per page once changed. When you go to watch the active pods and you change the number of rows after you enter a pod detail for example and you are back the number of rows is set back to the default of 10 rows.
d) Externalize the file that sync external repository inside the container. At the moment we have mpp.rsync and csa.rsync inside the mpp and the csa engine images that synchronize external customized files before starting the product from the /var/vols/itom/hcm/sync directory. This files (mpp.rsync and csa.rsync) includes only some directories and not others that in case of customer improvements needs to be synchronized as well, so I have created a new image only to change this two files. If these files are themselves external we do not need to change anything and we preserve the standard image for a future migration. In any case you have to mount the permament volumes before starting the product and this is only a reference file for the synchronization, so it will be an easy improvement with a great and powerful result.
e) Use the same concept also for IDM, creating an idm.rsync and put it externally as well, normally also idm is something that needs to be customized.
f) I have seen that the container are not synchronized in time with the host, they are in the UTC time zone and I have not seen any possibility to customize it. I think it could be a problem and I would like to avoid to customize all the images only to set the correct time zone. It would be a great improvement to have a global parameter to set the right time zone for all the containers.
g) At the moment there are a couple of bugs, at least in my installation. CSA 4.90 REST API "Add and Update component property" fails and practically no one CSA flows is able to complete as soon as it updates a component property, practically 99% of the csa flows. The second one is that with the integration of an external AD I am not able to login into the mpp portal from an AD user, also if it is found in the Lookup test, it works from the CSA admin console. There are two support tickets opened. Using a CSA 4.80 instance configured exactly the same with the same AD it works, the only difference is related to the DB, CSA 4.80 is installed in MSSQL, CSA 4.90 in the internal Postgres. Just ot be precise.
h) It would be nice to have a more detailed documentation about the architecture and how the entire system is connected and is working. This could be helpful in case of debugging problems. There are so many knowledge across how the architecture is built, products, docker, kubernetes, that as first approach is not easy, also if you have kubernetes and docker experience, because it is a customized installation and not a normal installation, so the solutions is different from a normal kubernetes standard installation and debugging problems (practically our daily job) is not easy.
e) CAC is mandatory for my customer to use this solution, please accelerate the integration, at the moment it is not yet supported.
I hope these few points could be helpful in some way, I am really enjoing the solutions and would really like to bring it into a production environment asap.
Thank you Mirco for your vlauable feedback.
When it comes to the out of the box capsules (eg. AWS, Microsoft Azure) what is your general feedback/feeling on the capsules released with HCM 2017.08 ?