This explains how to upgrade a cluster by scheduling the reboot and synchronizing time between both the nodes. If the devices are not part of a Chassis Cluster, refer to KB19925 - How to schedule upgrade reboot on SRX devices [juniper.net] .
user@srx> show system storage | match cf
user@srx>
show system storage | match cf
user@srx> request system storage cleanup
request system storage cleanup
user@srx> ftp <ip-address-of-local-ftp-server>
login using user credentials ftp> lcd /var/tmp ftp> bin ftp> get <junos-image-name> # i.e. junos-srxsme-10.2R3.10-domestic.tgz ftp> bye
user@srx> request system software add no-copy /var/tmp/ <junos-image-name>
request system software add no-copy /var/tmp/
<junos-image-name>
user@srx> request system reboot at 23:59
request system reboot at 23:59
user@srx> file copy /var/tmp/ <junos-image-name> or shell % rcp -T /cf/var/tmp/ <junos-image-name>
user@srx> request routing-engine login node 1