Hi,
I want to keep my sql DB synced with ELK.
I'm wondering, are there any advantages using jdbc scheduler against native linux cron?
thanks.
If you use the built-in scheduler you don't risk overlapping executions, i.e. that execution 2 starts before execution 1 has completed. Also, I don't think Logstash shuts down after the first run if you don't have a schedule set, so how would you know when to shut down?
You probably right with execution overlapping, though it's not the case for me. I plan to sync it once in 1h, each sync takes less than 1 minute. Regarding Logstash shutdown, I'm not quite understand. According to documentation " If no schedule is given, then the statement is run exactly once", and that's what I see.
My main concern is not to be dependent on connectivity to the ELK server.
You probably right with execution overlapping, though it's not the case for me. I plan to sync it once in 1h, each sync takes less than 1 minute.
In the normal case, yes. Do not assume that such will always be the case. I'm sure there are many pathological cases where a sync could take longer than that. What if the database server is down? What if you your switch or router breaks down and slows all network traffic to a crawl? What if both things happen at the same time? Unless you want to find out what happens I suggest you choose a resilient design.
Regarding Logstash shutdown, I'm not quite understand. According to documentation " If no schedule is given, then the statement is run exactly once", and that's what I see.
Yes, it's run only once but does it shut down Logstash after that sync? If not, you're going to start an extra Logstash instance every hour.
Yes, it's run only once but does it shut down Logstash after that sync? If not, you're going to start an extra Logstash instance every hour.
I guess so, I get this print at the end - Pipeline main has been shutdown
Anyway, @magnusbaeck - thanks for the clarification.