Ray, This looks odd. Sure, data has to be available for you after the restart, it's a basic capability. Then I tend to agree with the rest of the folks that you might haven encountered the issue mentioned by Yakov. To to be 100% sure about that please do the following:
* Clean "work/db" directory completely to avoid side effects. * Run the app that preloads data. Copy the generated logs somewhere. * Restart the cluster and start the second app. * Compare the first and second logs checking that the "page store" and WAL directories are identical in both files. * If the directories vary then you hit issue [1] and let us know if this the case. We have to come up how to solve this automatically. As a workaround try to send IgniteConfiguration.setConsistentId() to a predefined value. [1] https://issues.apache.org/jira/browse/IGNITE-6285 -- Denis On Fri, Sep 15, 2017 at 6:52 PM, Ray <ray...@cisco.com> wrote: > Yes, I tried to work with the data after restarting. > As I stated in the original post, I can't see any cache after I restart the > cluster. > So I called > https://apacheignite.readme.io/docs/rest-api#section-get-or-create-cache > API > with the exact cache name before restarting as parameter. > Then I try to run a sql query with that cache name and I got an empty > cache. > If the data loading from disk to RAM is transparent, I should see the data > before restarting, right? > > I agree with you Dmitriy, according to the debug log, the WAL folder is > resolved. > And I can confirm there're several WAL logs under that folder. > > > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >
