Create a Post
cancel
Showing results for 
Search instead for 
Did you mean: 
yishola
Participant
Jump to solution

R80.10 -> R80.20/30 Management upgrade issues

Hi There,

I've tried various upgrade paths for my VM Management server (R80.10 take 462) to R80.20 or R80.30 without success. I've increased the disk space and extended existing space with lvm_manager - still no joys. Tried cli and cpuse and the errors are always about insufficient disk space. I seem to have a lot of space.

Tried migrate export and space issue persists. Tried snapshot and though system says I need 9gb for snapshot (and I have 33gb free), snapshot is unsuccessful.

What I am looking for is a process by which I can upgrade the server without CheckPoint snapshot or backup. I can use VM Snapshot as fallback in case I need to.

LVM overview
============
                     Size(GB)     Used(GB)       Configurable    Description
lv_current  20                  9                    yes                     Check Point OS and products
lv_log          20                15                   yes                     Logs volume
upgrade     22                N/A                 no                      Reserved for version upgrade
swap           5                 N/A                 no                      Swap volume size
free             33               N/A                 no                      Unused space
------- ----
total 100 N/A no Total size

 

 

0 Kudos
1 Solution

Accepted Solutions
yishola
Participant

I resolved the issue and wanted to record what I did as it might help someone else:

1. Removed Endpoint Policy Manager blade from the FW Manager

2. Cleaned up my /boot filesystem to free up space

That was it. Upgrade was successful and I have gone as far as R80.40.

View solution in original post

0 Kudos
11 Replies
Dov_Fraivert
Employee
Employee

Hi Yishola,

Regarding the issue with the snapshot, what do you mean by "snapshot is unsuccessful"?
Does it give you an error?

Regards,
Dov

0 Kudos
yishola
Participant

The snapshot wasn't completing because of space issues. But as can be seen, I have loads of space.

Thanks

0 Kudos
Dov_Fraivert
Employee
Employee

Which error message did you receive?

0 Kudos
Tommy_Forrest
Advisor

Before upgrading my production machines to 80.30 I took backups of them and restored them into VM's using VMWare's Workstation product.

 

The first few builds I did, I did not create large enough disk volumes.  This was due to my workstation not having a ton of disk available.

I too experienced your issues later on (especially around the snapshot failures).  And using lvm_manager to extend the root volume would render the VM unbootable.  Every time.

In talking with TAC it was decided to just blow the system away and add more disk.  Which was frustrating.

Not apples-to-apples, I know.  But just wanted to note you're not alone with the snapshot issues when having plenty of disk space available and the upgrade not seeing it.

0 Kudos
yishola
Participant

Thanks for that.

 

I think I'll just have to start from scratch - or wait to see whether R80.40 plays ball.

 

 

0 Kudos
German_Cruz
Participant

with the lvm manager reduce the Logs volume 10 GB and it worked, I could already do the update

0 Kudos
yishola
Participant

This didn't work for me

0 Kudos
PhoneBoy
Admin
Admin
To leverage the xfs filesystem (which has better performance), it's actually better to do a migrate export/import from R80.10 to a freshly installed VM with R80.30.
0 Kudos
yishola
Participant

I started the exercise with a view to use migrate export/import.

./migrate export /tmp/gaia-r80-mgmt-2019-12-09.tgz

However the export fails each time with the following message:

Copying required files...
Execution finished with errors. See log file '/opt/CPshrd-R80/log/migrate-2019.12.09_12.32.06.log' for further details

Reading the log file doesn't particularly make any issues obvious, but I see the following lines (which still doesn't add up):

[9 Dec 12:35:04] .--> NoSpaceNotifierCleaner::exec
[9 Dec 12:35:04] [NoSpaceNotifierCleaner::exec] Starting notificator about lack of space during export
[9 Dec 12:35:04] ..--> GetConfig
[9 Dec 12:35:04] ..<-- GetConfig
[9 Dec 12:35:04] [NoSpaceNotifierCleaner::exec] Indication about gzip execution failure is: 0
[9 Dec 12:35:04] ..--> GetFreeDiskSpace
[9 Dec 12:35:04] [GetFreeDiskSpace] Directory '/opt/CPsuite-R80/fw1/tmp/migrate/' has 2649513984 bytes of free space
[9 Dec 12:35:04] ..<-- GetFreeDiskSpace
[9 Dec 12:35:04] ..--> UpgradeMacroReplacer::Instance
[9 Dec 12:35:04] ..<-- UpgradeMacroReplacer::Instance
[9 Dec 12:35:04] ..--> CanonicalizePath
[9 Dec 12:35:04] [CanonicalizePath] Canonicalizing path '/tmp/'
[9 Dec 12:35:04] [CanonicalizePath] Resulting path: '/tmp/'
[9 Dec 12:35:04] ..<-- CanonicalizePath
[9 Dec 12:35:04] ..--> GetFreeDiskSpace
[9 Dec 12:35:04] [GetFreeDiskSpace] Directory '/tmp/' has 10832781312 bytes of free space
[9 Dec 12:35:04] ..<-- GetFreeDiskSpace
[9 Dec 12:35:04] .<-- NoSpaceNotifierCleaner::exec
[9 Dec 12:35:04] .--> DirCleaner::exec
[9 Dec 12:35:04] [DirCleaner::exec] Going to remove directory '/opt/CPsuite-R80/fw1/tmp/migrate/'
[9 Dec 12:35:05] .<-- DirCleaner::exec
[9 Dec 12:35:05] [CleanupManager::DoCleanup] Completed the cleanup
[9 Dec 12:35:05] <-- CleanupManager::DoCleanup

 

0 Kudos
PhoneBoy
Admin
Admin
Best to engage with the TAC on this.
0 Kudos
yishola
Participant

I resolved the issue and wanted to record what I did as it might help someone else:

1. Removed Endpoint Policy Manager blade from the FW Manager

2. Cleaned up my /boot filesystem to free up space

That was it. Upgrade was successful and I have gone as far as R80.40.

0 Kudos

Leaderboard

Epsum factorial non deposit quid pro quo hic escorol.

Upcoming Events

    CheckMates Events