Solid State Drive (SSD) lifetime on QFX5100 platform might be shorten due to a software issue. In Virtual Chassis environment with Solid State Drive (SSD) installed on the Linecard side, the sdk-vmmd (app-engine-virtual-machine management service) should not try to establish the retry connection to SNMPD when it is running in line card because SNMP is not running in line cards members. A debug logging might be written to the SSD card frequently due to repetitive retry connection. Therefore, the SSD lifetime might be shorten due to the exceeded amount of read/write. Due to this issue, the SSD card failure might be observed. This issue might be seen if the following conditions are met: * Virtual Chassis platforms * SSD is installed on the line card members The software issue is tracked by PR1393044. https://prsearch.juniper.net/InfoCenter/index?page=prcontent&id=PR1393044
When the issue happens, high write I/O might be observed frequently: root@qfx5100> start shell % top -m io -o write -s 3 -n -d 20 PID USERNAME VCSW IVCSW READ WRITE FAULT TOTAL PERCENT COMMAND 1828 root 1232 17 0 1232 0 1232 100.00% sdk-vmmd <<<<<<<<<< High WRITE (1232) at the short intervals (-s 3 = 3 sec).. 23739 engineering 0 0 0 0 0 0 0.00% csh The SSD card failure might also be seen. The switches may crash during runtime and system not able to complete boot up. The following errors may observe in console logs during boot up: ata5: COMRESET failed (errno=-16) ata5: COMRESET failed (errno=-16) ata5: COMRESET failed (errno=-16) ata5: COMRESET failed (errno=-16) ata5: reset failed, giving up Unpacking root filesystem.... 873214 blocks Mounted root filesystem on /dev/ram0 Switching to newStarting udev: [ OK ] ERROR: expected SATA disks not found or not disk type. ERROR: Only found /dev/sda. ERROR: System cannot proceed without the disks #################################################### ERROR: Halting the system .... ####################################################