Hi Max, Thanks for getting back to me always. Comments inline!
> On Nov 9, 2017, at 16:58, Max Neunhöffer <[email protected]> wrote: > > Hi John, > > No golang setup is needed to run the ArangoDB starter `arangodb`. You can > simply use the supplied executable. I did discover this! I think I did see some old information somewhere on Github. > > By now, all binary packages we release should contain the starter executable. > If this is not the case, this is a bug. I am no expert in Mac packaging and > have no clue how the situation is with homebrew or the app bundle we ship. We > need the Mac experts to look into this, I will alert Frank. In any case, you > can always just download the latest starter executable from > https://github.com/arangodb-helper/arangodb just see the releases thete. It is missing from the homebrew version but definitely in the direct download ArangoDB-CLI.app After inspecting the app bundle contents, I could see quickly how everything was laid out there, and it makes great sense for easy upgrades and deployments. Quite nice in fact! > > As to the name: our idea is that the arangodb starter tool is the primary > tool to start ArangoDB in all configurations: single server, cluster and the > upcoming single server resilience config. The user should either use a > package in which ArangoDB is started automatically, or use the starter (or > some orchestration solution like mesos or kubernetes). That makes sense! Especially after using it successfully. > > In the standard configuration the starter starts a cluster with a coordinator > on each machine, by default on port 8529. You can then contact any of these > endpoints to talk to the cluster. This at first concerns endpoints. You can > use ip addresses or host names as you like. > > However, your DNS configuration is completely outside the scope of the > starter and ArangoDB. You have to manage this by other means. It turns out, so simple and impressive. I could start the cluster, join any node at any DNS entry for it, and raise the cluster. I honestly do not know how chatty this will make my network, but I intend to put an instance up on each node in a Jenkins CI. Then things that run on each Jenkins CI node can try to reach the first available ArangoDB cluster node. They start with localhost, then query all the IPs/URLs of the nodes in Jenkins, until one is reached to do database things with. In theory, this is simple, reachable and fast, and easy to keep up and available. One of my use cases is a simple pool of resources in-use. Each Jenkins CI node checks the pool first. If resource is not in the pool, the Jenkins CI node claims the resource by adding to the pool. If resource is in the pool already, Jenkins CI node must choose another resource. > > I hope this answers your questions. I will alert the Macexperts about your > other questions regarding homebrew. It does! > > Cheers, Max > > -- > You received this message because you are subscribed to a topic in the Google > Groups "ArangoDB" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/arangodb/HGbsnKK63i0/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected]. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "ArangoDB" 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.
