Fortigate process " wad" consuming 62% of memory.

[plindgren]

i get the " CFG_CMDBAPI_ERR" when i try to make changes on my fortigate. which is other than that operational. I tired the command " diag test application ipsmonitor 99" but it did not work. So i used the command " diag sys top 1" to see what was hogging all that memory. And i found a process named " wad" that uses 62% of the memory. However this machine is in production and i dont know what the process does and i cant seem to find it anywhere. Can i kill it? What does it do? Is there a process reference for fortios out there somewhere?

[figge]

Memory use is still increasing day-by-day, even after the upgrade to 5.6.8.

The increase is slower, 1% per day approximately, it will be interesting to see

if/when we reach 80%.

[itismo]

Upgraded and no issues yet :)

[mike_dp]

we had the same issue with 6.0.0 to 6.0.4 and it's now fixed when we upgraded to 6.0.5.

[Fresh]

Are you sure is already fixed in that version, cause I have 6.0.4 and I still with the same issue,

[nikade87]

I am running v6.2.0 build0866 (GA) on 2st Fortigate 200E in a HA-cluster and I am also seeing this.

Yesterday it was "ipsmonitor" which had 3 processes using about 20% each, today it was "wad" using about 40%.

I have created a ticket with the Fortinet support hoping they have a solution since it is causing a lot of trouble in our main-office where we have this setup.

[nikade87]

Support told me this is a known issue in 6.2.0 and recommended me to upgrade to 6.2.1 to resolve the memory-issue.

[nikade87]

After upgrading to 6.2.1 the memory-issue has gone away, but a new problem has been detected. When I have the "application control default"-policy enabled on rules with a lot of traffic and sessions the CPU is spiking at 99% on one of the cores, this wasn't the case on 6.2.0. Support is working on resolving it so I am crossing my fingers.

[Lufa]

Hello guys! Any updates on this? We are facing both lack of RAM and CPU peaks on 6.2.1.

[nikade87]

Yes, I think we have found a bug in the wad. Also we got a new version of the ipsengine-database and since this we havent had any issues with either CPU or RAM-usage.

Regarding the bug their developers are working on it, it caused the wad process to crash frequently before we got the new version of ipsengine-database.

[Lufa]

nikade87 wrote:

Also we got a new version of the ipsengine-database 

Do you mean regular updates, or some special update? How can other people achieve this?