I seem to remember a configurable thread limit per ultragrep process. This appears to be no longer the case. Ultragrep is crushing the logs servers with a thread per log file searched. There needs to be some cap. We are delaying/losing logs, and paging on-call because of it, on a daily basis due to single ultragrep runs.
I seem to remember a configurable thread limit per ultragrep process. This appears to be no longer the case. Ultragrep is crushing the logs servers with a thread per log file searched. There needs to be some cap. We are delaying/losing logs, and paging on-call because of it, on a daily basis due to single ultragrep runs.