> On Mar 6, 2023, at 12:03, <jonmcalexan...@wellsfargo.com.invalid> > <jonmcalexan...@wellsfargo.com.INVALID> wrote: > > I believe we may have figured out what was causing the issue, but not why it > was causing it. > > In the mod-proxy-http configuration for the balancer, we had been using the > fully qualified host name in the balancer url. When we switched to IP > address, things got tremendously faster to the point we aren't seeing the > health check fail any longer. What is perplexing is why removing a simple DNS > lookup by going from host name to IP address had such a dramatic affect. Is > this something that adding ProxyHost and ProxyPort to the connector would > possibly cure? We have not tried that currently.
Once upon a time (before I retired), we had slow DNS lookups due to a misconfigured /etc/resolv.conf file still including a disabled DNS box as the first in the list. Might want to check the DNS configurations on the systems of interest and insure that all listed DNS boxes are functional. - Chuck --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
