We’d need to see more of the stack trace and see your configuration before this could be diagnosed.
It looks like a node abruptly left the cluster. If you’re losing data it seems like you’ve not configured any backups. > On 10 Mar 2021, at 16:53, Josh Katz <[email protected]> > wrote: > > After we load 35 million records from the database into the cluster. When we > query it using a scan query we get the following SEVERE error, > After that the node are restarting and the caches all show 0 keys and we have > to restart the initial load. > Is there a way to salvage the data so we don’t have to reload every time? For > example, if we add persistence at this stage the initial load will be slower > correct? > How can we avoid seeing this error? “[SEVERE] Failed to process selector key” > > <image001.png> > > Thanks, > Josh > > > Please follow the hyperlink to important disclosures. > https://www.dodgeandcox.com/disclosures/email_disclosure_funds.html > <https://www.dodgeandcox.com/disclosures/email_disclosure_funds.html>
