MTA is skipping email messages due to disk space?

[Ivan_Lupasco]

Hello all!

Cluster standalone environment, 2x r80.10 5600 appliances. I am deploying MTA in a company. All config done by sk109699. Now I see that testing messages are coming to recipients without Checkpoint sign(box is checked). Also in logs I see skipped email status of testing messages (screenshot in attachment). In additional log info I see that reason is reaching the disk space limit, but my var/log is used only 55%. Where can be the problem?

‌ ‌

[Norbert_Bohusch]

Can you post the configuration of the MTA (cluster object -> Mail Transfer Object -> Advanced Settings). 

So we see how much maximum disk usage you have configured (% of storage or MB).

Then also an output of "df -h" from gateway.

[Ivan_Lupasco]

Thank you for your reply!

The requested information is in attachment:

[Norbert_Bohusch]

You have configured a "maximum usage of 40%" and the actual usage is 55% and you are wondering that it is not working?

Maybe you misunderstood the setting, but the "maximum usage" defines after which disk usage the MTA is stopping to work.

So you should put it to 60% or 70% (default value) and it will start to work again.

[Ivan_Lupasco]

I don't remember that I have changed this value. I left it as it was.

Anyway I thought that this value defines the maximum amount of disk usage by MTA only and not by other processes. Can you please point me the documented explanation of this parameter? The second option you can choose is MB value(1024-default), and in conformity to your words it should say: after the use of 1024 MB (from all blades) MTA should stop working. I don't remember anytime when var/log was only 1024 MB used. May be I am misunderstanding... After all thank you so much for this hint!

The point is maybe I should then change this value to 20?. In this case I should have 20 from MTA and 55 from other blades and var/log will be less then 80% used.

[Norbert_Bohusch · Accepted solution]

I don't know why this is different, but as far as I understand it, the %-value looks at whole usage of /var/log partition, but the MB-value only looks on free disk space. Else the default values make no sense.

so you should enlarge the %-value

[Ivan_Lupasco]

Just have looked in smartconsole help. I think I should make it less than 40%. Value defines the Percentage of free hard disk space that the MTA can use. Please correct me if I am wrong.

[Norbert_Bohusch]

I don’t think so, but just try it. As it is not working at the moment, you can’t break anything 😉

[Ivan_Lupasco]

Ok, I will definitely try, and update here soon!

[Ivan_Lupasco]

Hello Norbert! I guess you were right! I have changed the value to 20%- and.... nothing)) Then changed to 70, and I think it started to work. At least in the MTA logs I don't see Email status of "skipped". Now it is "bounced" think that is because of sender is root@localhost - test message.

So thank you again!

[Chris_Atkinson]

Some useful content is available in the following related thread:

https://community.checkpoint.com/message/37034-we-have-threat-emulation-issues-due-to-disk-space