A vendor has provided us with a chef cookbook for their product. This cookbook uses attributes that are set to host names of nodes to install software on specific nodes and to configuration communication between the nodes. For instance, there is an attribute for the zookeeper hosts and an attribute for a task manager hosts, along with the attribute for the overall node hostnames for their core product. The core product needs to know what the zookeeper hosts are, and what the task manager hosts are. What complicates this further is that the zookeeper and the task manager need to run on the same nodes as their core product although zookeeper, task manager, and their other tools do not need to all run on the same nodes. Zookeeper would run on three nodes, the task manager would run on two nodes that zookeeper is not running on, and so on with the other applications.
What it boils down to is that the cookbook expects that all the hostnames are defined for specific roles before executing. I thought about setting a gobal variable for these hosts so that when a zookeeper host comes up it registers itself to that global varaible, when the next zookeeper host comes up it adds itself to that variable so that end state that variable has all the zookeepers. But that does not solve the problem that the first node to come up will go through the chef run before all the zookeepers are known. Hence the reason to want to stop the nodes before the hostup process so that all this information can be collected into the global variables and used as override attributes. Perhaps there is a better way to do this? On Wednesday, June 17, 2015 at 1:32:51 PM UTC-4, Sebastian Stadil wrote: > > Can you tell us a bit about the problem behind the question? There might > be a better solution than the proposed. > > On Wed, Jun 17, 2015 at 10:19 AM Dann S Washko <[email protected] > <javascript:>> wrote: > >> Is it possible to pause the execution of a node say before host, and do >> this for each node in the farm, until the final node is brought up, perform >> an orchestrated action, and resume the execution process across all nodes? >> >> -- >> You received this message because you are subscribed to the Google Groups >> "scalr-discuss" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> For more options, visit https://groups.google.com/d/optout. >> > -- You received this message because you are subscribed to the Google Groups "scalr-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
