This is a very common question that we get asked by both new and existing customers who want to get the best outcome and results from MapR Support. The following are some of the steps I would recommend to speed things up, based on the criticality of the issues.
Let's start with criticality first. MapR has labeled P1, P2, P3, and P4 as criticalities, defining what they mean in the context of development and production environments. For example, a cluster down or job failure can be seen as P1 in production but not in a development environment. To understand the SLAs and definitions of criticalities, please refer to the service offerings link on our website.
Let's now focus on creating a case. The simplest way is to become a customer and enable your login for Support Portal access. This is usually done at the time of closing by working with the account team or sending an email to firstname.lastname@example.org. Once you have access to the Support Portal, you gain the advantage of setting priority, escalating when necessary, looking at the existing bugs that are released as patches, and searching for knowledge articles. Using email@example.com doesn't give you these privileges; plus, the email will be decommissioned soon. Picking the right release and category for the issue is the next most important step.
Let's discuss the what/when/where/who/how of the issue at hand. As you know, getting quick answers depends on explaining what, when, where, and how the problem surfaced. For example, stating that an issue happened today (when) with some frequency in a production environment (where) because of change (what) that was introduced by an admin, user, or app developer (who) with an impact goes a long way in solving the problem.
Finally, let's go over what to package and send. This is likely the last step in the journey. MapR engineers will need logs, core files, job failure exceptions, and diagnostics to determine the failure. We don't keep your logs or information about your environment in our systems. To speed things up, send us the following to help us quickly resolve the issue:
-Diagnostics is collected by executing the support dump utility:
–If not related to MapReduce, you can exclude Hadoop logs, etc.
–Support dumps should be run on nodes that are being affected:
Sending MapR log files, ecosystem logs, core files, sysinfo, and cluster configuration helps to resolve issues rapidly. The Support Portal has a limitation of file size to attach. As part of enabling your access to the Support Portal, you will be given access to a secure ftp site with username and password for uploads.