2009.8.6 SOL7036: The Linux uptime counter wraps after 497 days

SOL7036: The Linux uptime counter wraps after 497 days

---

Known Issue
Updated: 7/21/09 11:05 AM

This is the result of a known issue, which was discovered in 1998 and still exists in the Linux 2.4 kernel. The kernel computes the system uptime based on the internal jiffies counter, which counts the time since boot in units of 10 milliseconds, or jiffies. The counter is a 32-bit counter, which has a maximum value of 2^32, or 4,294,967,296. When the counter reaches this value (after 497 days, 2 hours, 27 minutes, and 53 seconds, or approximately 16 months), it wraps back around to zero and continues to increment.

F5 Networks Product Development is tracking this Linux issue as CR74550. Currently, no Linux kernel patch is available to resolve this issue.

When the counter wraps, the following side effects may be observed:

• The ps (Process Status) command may report incorrect TIME values for daemons that were running at the time of the counter wrap
• Some processes that depend on accurate elapsed time calculations may be adversely affected, as detailed in the following Solutions:

SOL7071: SCCP kernel driver i2c read failure

SOL8087: SCCP kernel driver timer wrap may cause system component health misreadings (FirePass only)

SOL9679: The lacpd daemon stops sending LACP messages after 497 day linux uptime wraparound

SOL9683: The gtmd, tmm, or bcm56xxd daemons may crash after 497 day linux uptime wraparound

SOL10311: The performance graphs no longer display data after 497 day linux uptime wraparound

Note: These are the only known side effects of the uptime counter wrapping. The issues documented in these Solutions have been patched as noted therein. If you encounter issues that seem related but are not documented here, contact F5 Networks Technical Support.

Workaround

You can work around most issues created by the wrapping of the uptime counter by rebooting the system. In some cases, further remedial steps may be necessary. Refer to the solutions above for specifics.

You can avoid this issue by rebooting the system prior to the 497 day counter wrap. To determine how long the system has been up, use the Linux uptime command. The uptime command produces output that appears similar to the following example:

19:52:48 up 20 days, 9:24, 1 user, load average: 0.09, 0.05, 0.11