What maximum single disk size is supported by R80.20 security management server? Or SMS upgrade to R80.20 with disk growth.

[EVSolovyev]

Hi all.

Today I hava a problem. I must upgrade R80.10 security management server to R80.20. I choose fresh install + migrate export/import, because CPSUE upgrade does not change linux kernel to 3.10 (as I understand). And during my works I see, that customer has logs only for 4 months (about 200 GB). Customer tell me, that he need to store logs as long, as it possiable and he has 4 TB virtual disk (SMS is virtual open server on vmware).

I use this ISO image for new management server:

Support, Support Requests, Training, Documentation, and Knowledge base for Check Point products and ... 

I use this migration tool and I think, that my problem can be interesting to all, whose SMS below R80.20.

Support, Support Requests, Training, Documentation, and Knowledge base for Check Point products and ... 

I think, that 4 TB disk is too big, but I see in to release notes:

TAGS ARE REQUIRED --> Check Point Software Technologies: Download Center 

What is more than 2 TB I think? I know, that Check Point based on RHEL. So I go to RHEL and what I see?

https://access.redhat.com/solutions/1532 

So all is OK and 4 TB is more, than 2 TB I tnink. And we started.

Here is my disk scale:

But at the next step (4th) system freezes during the entering password. I tried 3 times (reboot VM and start again). But I couldn't do it.

After that I go to lab and create VM with 2 TB storage (it's my maximum lab disk capacity). And I see no problem. I can do migrate import and all is fine. SMS works.

Specify exact value of single disk maximum capasity in release notes, but not "More than 2TB support per a single storage device", please.

[G_W_Albrecht]

This must be caused by something and i would suggest to involve TAC ! When looking at the Smart-1 models, you find that storage on Smart-1 505 is 2 x 4TB, on 5050 4 x 4TB and on 5150 up to 12 x 4TB ! So that has to work...

[G_W_Albrecht]

I just have heard that this could be a VMFS issue - two possible causes:

- VMFS is an older version that will not work with 4TB

- VM has been e.g. 500GB, got enlarged to 4TB, but will not work although VMFS is a newer one supporting that

[EVSolovyev]

Hi. That's interesting info. Thank you.

[G_W_Albrecht]

This also looks rather like a problem of VMFS as the installer itself neither crashes, nor stops the installation.

[EVSolovyev]

I think a problem is the disk > 2TB and not EFI. I'll test my thinks and write about it.

Last night I can't install SMS again and has no thinks what to go. But my collegue tell my this way and I think, that he's right.

[G_W_Albrecht]

Concerning disk controller type, you have no choice - but you know this documentation ? sk88520: Best Practices - Identity Awareness Large Scale Deployment

[EVSolovyev]

So. I had a test day today in my lab and what I see? SMS with single disk more than 2TB can be deploid only by web console of vmware.

I test it on my ESXi 6.0. And on monday will test in on ESXi 6.5.

If we create VM with disk 2 TB (2048 GB) or 4 TB (4096 GB) after succes installation and reboot I had no system. .Only that:

or that:

If I get disk 1997 GB, 3 TB, 5 TB or 6 TB - I install system, reboot, migrate import and all is fine.

If you install system with single disk less 2TB - there is no f****g magic.

If you need single disk size more 2 TB, you need some magic. I can deploy SMS only with that options:

1. You MUST use ONLY web vmware client. In programm client I can't install SM Swith disk more 2 TB.

2. OS - Linux RHEL 7.

3. ISO image must been import to datastore and connected to VM at first power on.

4. SCSI controller - LSI Logic SAS.

5. I had no space for thick disk - so I test on thin disk. In product I think disk must be thick.

6. Boot options must be "EFI in legacy compability mode"

Complete options must be like this:

I don't know why. And I don't know is it VMware bug, or Check Point bug. But I think it is bug.

P.S.

With my case.

This does not work:

And this does not work:

This works:

For 5 and 6 TB I have no screens, but was tested and it boots. On VM with 5 TB customers import was migrated (in lab inftastructure) and I see no problem.

Hope it helps somebody, becouse fore me in was 2 nights and 2 days.

[Maarten_Sjouw]

On the last screens you show very small system discs for a MDS, what was the problem you meant in the disc size screens for SMS?

I would need at least 650GB for the system disc and the rest of a 4TB for /var/log

[EVSolovyev]

Hi. My stystem is not multidomain. I see no need to a big  root. Backup I think not need more 200 GB for me, because in my cases I dont see snapshot file more than 20 GB.

[EVSolovyev]

Today install R80.20 with 3 TB disk in product. Migrate import. See no problem.

[EVSolovyev]

The next bug, that I catch after migrate import on a cleen server, was a licence fail.

Both nodes of a cluster told me, that there is a licence problem. I got it after 2nd policy installation.

My FWs are directly connecter to Internet. And curls were aliable.

After reattaching licences all exept URL filtering went away except one - URL filtering.

At the end of my attampts to fix it - I reboot both nodes (first - active, wait, cluster becamer eady, after that I reboot active again (standby node became active)).

After that cluster get normal status.

SMS - R80.20, open server.

FW - cluster of 5800, R80.10 (may be would be updated to R80.20 too).