This article explains that configuring a small traceoptions file size may lead to scheduler slips and cause protocols like LDP to flap.

The customer committed BGP traceoptions, leading to a disruption in LDP.

Upon analyzing the logs, a scheduler slip was observed immediately after the commit, accompanied by a jsr_unreplicate message, indicating a potential issue with LDP.

The following trace configuration forces the file size to be extremely small and continuously rotates between the two trace files it generates, potentially keeping disk I/O excessively busy:

traceoptions {

  file bgp-log size 10k files 2;

  flag update detail;

}

The continuous slips suggest that the system or user is unable to obtain resources, meaning the kernel itself is not receiving CPU cycles. This typically occurs when other processes, such as extensive disk I/O from traceoptions, are consuming CPU cycles.

The work around is to use a large file size to reduce the file rotation.