topic Re: DB corruption after power failure? in Firewall and Security Management

DB corruption after power failure?

Nishad_Abooback — Mon, 14 Jan 2019 12:42:05 GMT

Hello friends,

we had a power issue in one of our sites due to UPS issue, checkpoint 5200 appliance running as a gateway rebooted, it came up without security policy , fw stat - showed as "default policy" and it didn't pass any traffic. It also lost the connection with mgmt server. Resetting SIC didn't help, as when I pushed policy I got error "load on module failed". I had to restore the gateway to factory defaults to get it back working. This was the second time we faced such an issue, earlier with a different site. Worked with TAC, they said some DB corruption and did the same (restored to factory defaults) as a fix. Is there a way we can avoid this kind of corruption ?

Thanks,

Nishad

Re: DB corruption after power failure?

Maarten_Sjouw — Mon, 14 Jan 2019 13:58:26 GMT

I thought this behavior was no longer happening, we had this type of corruption quite a lot with 2012 appliances (like 2200 and 4x00) in Asian and African countries where power outages are quite often happening. The problem occurs when a writew action is taking place during the power cut. Mostly the only way to recover is a reinstall, either from USB or from a snapshot.

When you're lucky you might be able to recover using fsck while booted in maintenance mode.

Re: DB corruption after power failure?

Kaspars_Zibarts — Mon, 14 Jan 2019 21:53:52 GMT

I agree. On such sites I with risk of power failure I would always have a snapshot on local and external disk. Then you're back in business fairly quickly - just boot using snapshot

Re: DB corruption after power failure?

Nishad_Abooback — Tue, 15 Jan 2019 09:39:12 GMT

Thank you!