Due to some reason, Junos upgrade failure might occur in the middle of the ISSU procedure. The warning: ISSU in progress error message is displayed and commit also fails. The device cannot be rebooted via the request system reboot command.
Customers experience an issue, when attempting a ISSU upgrade on an EX4200 virtual chassis. The system generates the warning: ISSU in progress error message:
{master:0}[edit] root# commit warning: ISSU in progress, commit disallowed {master:0}[edit] root# commit force warning: ISSU in progress, commit disallowed {master:0}[edit] root# exit Exiting configuration mode {master:0} root> request system reboot warning: ISSU in progress {master:0} root>
This issue is due to a flag been set, when performing the ISSU upgrade. The flag bit will not get cleared; so ISSU experienced an unexpected failure. The customer can check the flag via the root shell by using the following command; a non-zero value indicates that the ISSU flag has been set:
root@:RE:0% sysctl hw.re.issu_state hw.re.issu_state: 64
root@:RE:0% nvram getenv hw.ore.issustate hw.ore.issustate=64
To resolve this issue, you have to clear the flag via the following shell command:
root@:RE:0& nvram unsetenv hw.ore.issustate
root@:RE:0% nvram setenv hw.ore.issustate 0
root@:RE:0& nvram unsetenv hw.lc.issuboot
root@:RE:0% sysctl -w hw.re.issu_state=0 hw.re.issu_state: 64 -> 0
{master:0}[edit] root# commit warning: ISSU in progress, commit disallowed {master:0}[edit] root# exit Exiting configuration mode {master:0} root> request system reboot warning: ISSU in progress {master:0} root> start shell user root root@:RE:0%sysctl hw.re.issu_state hw.re.issu_state: 64 root@:RE:0% sysctl -w hw.re.issu_state=0 hw.re.issu_state: 64 -> 0 root@:RE:0% nvram unsetenv hw.ore.issustate root@:RE:0% nvram getenv hw.ore.issustate nvram: variable hw.ore.issustate not found root@:RE:0% cli {master:0} root> configure Entering configuration mode {master:0}[edit] root# commit configuration check succeeds commit complete {master:0}[edit] root#