SSH and GUI gone bad; The rest works fine

[gilfalko]

I have 2 clustered 200B units with 5.0 FW installed. Last night for some reason both the GUI and SSH completely stopped working. The WEB portal just hangs before the login screen and once you punch in your password via SSH it also hangs. The funny thing is everything else still works just fine. VPNs, incoming\outgoing traffic etc all is good. Only the management part is completely broken. I might as well add that this happened about 2 days right after I registered both units which are currently out of warranty plus added them to the cloud. Has this ever happened to anyone?

[Dave_Hall]

Only with the smaller fgts that go into conserve mode and/or the log device is having issues. Rebooting the affected units temporary fixes the conserve mode problem and reformatting the logging device helps with the logging issues. I assume you have logging enabled? -- it should hopefully clue you in on what' s happening. Not in front of a fgt device, so not sure of the syntax for using TOP to check the system resources.

[netmin]

Just recently we had attempted to reconfigure the GUI dashboards via GUI and CLI (5.2) for the reason mentioned below. The primary FGT didn' t really like this and GUI went from slow to non-responding, CLI didn' t really want to talk to us as well (though login was still possible). At this time, cmbdsvr process was consuming 100% of one cpu core, but cpu consumption went back to normal after a minute or two. Later, GUI went down completely. Traffic appeared to flow as normal. So we logged into the secondary, via dedicated mgmt port/ip and GUI/CLI appeared normal there. We forced it into the active role (diag sys ha set-as-master enable). From then on, it became slower as well, when trying to modify some setting (did not sync to the first FGT). We then accessed the first (now passive) FGT via the HA link (exec ha manage 1) and rebooted it. This brought speed back to normal on both FGTs. Later on, failback via diag sys ha set-as-master disable and diag sys ha reset-uptime to switch the roles back. Mentioned in another post: since upgrade to 5.2 some GUI widgets did not show any traffic (others did) and although both devices were rebooted more than once since, there seemed not to be any easy way to change it. After this ' hang condition' the FGT decided to NOW show the traffic - so any previous dashboard configuration (from 5.0.x) appeared to cause this - which was now cleared. Yes, the configuration was checked after firmware upgrade and nothing obviously wrong was observed.