IBM Power Connect, learn, share, and engage with IBM Power. Join / Log in
Every 4.0s: MONHmc mem Thu Nov 12 08:41:59 2020
KiB Mem : 32919864 total, 477068 free, 20001772 used, 12441024 buff/cache
Every 4.0s: MONHmc mem Thu Nov 12 08:41:59 2020KiB Mem : 32919864 total, 477068 free, 20001772 used, 12441024 buff/cache
I have several HMCs, my practice is:
1. Always use the dual-HMC environment, at least for production, to avoid such problems.
2. Old HMC hardware works not too good with new HMC software. It's obvious. it's officially not supported (?), but addding some extra OEM memory DIMMs may help.
3. If you "feel" that your HMC became slower - reboot it!
4. Configure the IMM interface at the HMC hardware server, it will help to reboot the hanged one.monhmc is the good tool, but I've practically never used it and don't worry about it.It's my practical view only.
3. If you "feel" that your HMC became slower - reboot it! Is my recommendation too, but sometimes no time to reboot when you detect the HMC is slow ... The tip of Steve Munday looks interesting with momhmc -r MemFernando
Fernando,Surely monhmc is a good advice, but in case you have no time to reboot - see "1": dual HMC.
Anyway it's a very smart idea, you may achieve better fault prediction if intergrate it with your monitoring solution (Tivoli, HP, Zabbix or anything else):
grant the necessary permission to the Monitoring, run the monhmc on schedule, alert on threshold.
Use ssh keys to login to the HMC without password.