Internal Storage Failing - Is this user replaceable?

[BrianT]

I have a Fortigate 60C that suddenly seems to no longer see its internal storage. In the " System Resources" widget in the Status portion of the Dashboard, the " Disk Usage" dial has suddenly gone missing. I have switched my logging over to upload to FortiCloud in real time for the time being, but I was having issues a bit earlier today trying to get the firewall to display logs from FortiCloud earlier so I' d like to get log to disk working again. Is the internal storage something that is user replaceable? This unit isn' t even a year old yet and I' d prefer to not have to deal with a refurb unit already since I' m sure that' s what will happen if I have the hardware serviced by FortiNet. Thanks in advance for any info anyone can provide. P.S. Seeing the following errors in the event log, internal storage is toast: 1 16:28:26 EXT3-fs error (device sd(8,4)): ext3_get_inode_loc: unable to read inode block - inode=162954, block=327684 2 16:28:26 EXT3-fs error (device sd(8,4)) in ext3_orphan_add: IO failure 3 16:28:26 EXT3-fs error (device sd(8,4)) in ext3_reserve_inode_write: IO failure 4 16:28:26 EXT3-fs error (devoc: unable to read inode block - inode=15, block=4

[AtiT]

Hi, It seems that the filesystem on the disk is corrupted or the disk is broken. Try to format the logdisk - probably reboot will be needed and some minutes downtime during the disk formatting. Firstly, I suggest to open a trouble ticket at fortinet tac - we had some fortigates with EXT3-fs error and they were sent to RMA.

[ShrewLWD]

Hey BrianT, Of my last 24 60Cs I purchased (last 3 months), 9 failed with that same thing. I' m convinced they have switched flash provider, and QA has taken a nose-dive. I' m not getting much response back from TAC or my SE, but its still a fairly new thing. Formatting the device or the logdisk will not solve it. Once you get those errors, RMA it out, it will continue to fail in the EXT3 and EXT2 areas.

[Warren_Olson_FTNT]

I have extensive 60C experience and can say we saw very few RMAs that weren' t caused by electrical/storm issues in areas they (for whatever reason) didnt plug into a UPS. If you' re doing a lot of local(disk) logging I could however see the flash eventually have issues. If you end up having to RMA the device I would recommend not logging locally to disk and pushing all logs to FortiAnalyzer/syslog server.

[ShrewLWD]

Hi Warren, We have standardized on 60Cs (over 150 now), but I can assure you, something has changed very recently. ALL of ours are plugged into UPS units. In one case, I had a 60C fail within 2 days (logging enabled), while another failed within a week with logging disabled. Others have been running 2 years, with full logging enabled. Without finding a root cause, we have had to disable all logging and rely on syslog data only. The only explanation at this point is either a change in flash manufacturer, or a batch with very poor QA. All the devices were fairly close in serial number range.

[Warren_Olson_FTNT]

It' s a good idea to disable logging on any of the SOHO models, which may be the default option in 5.x, it was definitely changed to default in 4.x. I would still probably open a support case and reference the serial number batch and have them investigate for you, no reason to keep replacing them which isn' t beneficial for either side.

[Sean_Toomey_FTNT]

Yes best practice is to disable disk logging on smaller models that do not have SSD. The reason why is that storage on the 60C is flash based and disk logging will result in a heavy read/write duty cycle and lessen the lifespan of the disk. This is why it will warn you in 5.x if you try to enable disk logging that this may happen. However, it is still fine to log to memory locally if you just need logs in the very short term local to the box. Beyond that, my recommendation is the same as Warren' s, please catalog all the serial numbers and information and open a TAC case so that we can properly investigate your situation and drive it to resolution. Cheers!

[dasilva13]

I have to agree with the other posts. I have about 30 60C out in the field and have the same problems. There is no rhyme or reason as to when they die but they do often. I have some with FortiCloud/FortiAnalyzer/local Syslog and some without logging altogether and have many issues with these small boxes. We have tried to stay away from them as much as possible and smallest we try to go now is 100d' s (even though the price point isn' t exactly for SMB)

[emnoc]

Same here ( FGT60C falling like rain drops when it comes to disks ). If logging is such a bad thing, why dos fortinet place logdisk in these units knowing that they can fail. Forticloud and FAZ are not the solutions ( for everybody ) & and specially when we have a disk inside of a unit that... pretty much useless. I personally think this is cop-out and fortinet selling us short imho. It liks having a OD button on your shifter but the mfg' er tells you can' t use it because of xyz ....

[Ralph1973]

Hello, I had the same issue with my privately purchased 60c, however hardware support had just stopped since 2 months, so I have a defective unit here and I am not able to replace the flash, since it sticks to the motherboard :-( Regards, Ralph Willemsen