Managing Core Dump Files

When a core dump occurs on a redundant SRP and you have the router configured to store network core dumps, the SRP that experiences the trouble retains the management Ethernet port to perform the core dump. This prevents the standby SRP from taking over operations until the core dump is complete.

When a router uses local NVS to store a core dump, the SRP does not need the management Ethernet port. However, because of the immense size of local core dump files, using NVS to store core dumps is not practical.

The SRP-120 available on the E120 router and the SRP-320 available on the E120 and E320 routers has a second NVS card which is dedicated to storing core dump files.

The core dump monitor eliminates the impact that core dumps may have on redundant routers by allowing you to manage core dump files in NVS. The core dump monitor allows you to automatically transfer core dump files from NVS to an FTP server location for storage. The core dump monitor can also automatically delete transferred core dump files.

The core dump monitor attempts to delete transferred files when all of the following conditions have been met:

• The router attempts to write a core dump file to NVS.
• NVS contains insufficient space to hold the new core dump file.
• The core dump files have already been transferred from NVS to an FTP server location using the automatic core dump monitor transfer process.

Only those core dump files that have already been transferred from NVS are considered for deletion. Of those, the oldest files are deleted first, and the router generates a log message for each core dump file it deletes.

Note: If the router NVS does not contain sufficient space to hold a new core dump file even after deleting all possible core dump files, the core dump fails and the router generates a log message for this condition.

Copyright © 2008, Juniper Networks, Inc. All rights reserved. Trademark Notice.