Date: Mon, 25 Apr 2022 05:12:29 UTC
On 24 Apr, Rene Ladan wrote: > On Mon, Apr 18, 2022 at 11:38:36PM +0000, Pau Amma wrote: >> On 2022-04-18 22:10, Mark Millard wrote: >> > Looking at the log shows other failures during 97% : >> >> [massive snip] >> >> > I suspect that the rest of the time was its very slow scanning >> > of the huge log file to report on the type of failure. This can >> > take 2+ hours of itself before the actual kill happens. I ran into this issue recently with a non-standard build of openoffice that produced multi-GB log files. The scan was taking > 24 hours. It might be more efficient to use a sed script that quits on first match instead of all of those greps. Something using hyperscan would probably be even better.