question

jonoci avatar image
jonoci asked

Cerbo GX data file storage is full - how to make space

The data file area on my cerbo is full - I have many nodes deployed and have hit the limit it seems. When looking with SSH at

/data/home/nodered/.node-red/.node-red# directory, there are many installed nodes that are not used - eg, home assistant, chatbot, @slack, etc - these do not appear in the node list in pallet manager in node red, so cannnot remove them from there. Is there a way to safely remove these (many) unused node:contrib nodes via SSH - root? there are 100's of MB of files like that that I do not seem to need.


storage
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

pwfarnell avatar image pwfarnell commented ·
Moved to modifications as it may get more traction there.
0 Likes 0 ·
4 Answers
pwfarnell avatar image
pwfarnell answered ·

Do either of these earlier questions help you out, I know they do not answer the specific question of deleting nodes, but may offer some thoughts on resizing space.

https://community.victronenergy.com/questions/131898/cerbo-gx-storage-full.html

https://community.victronenergy.com/questions/230999/root-filesystem-100-full-how-to-use-resize2fs.html

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

jonoci avatar image jonoci commented ·
Thanks for this, but I did find these already and unfortunately are of no help - I cannot even delete nodes from my dashboard - it won't allow a deploy - comes back say no space...A catch 22, needs space to make space...so it has to be done at root SSH linux level..
0 Likes 0 ·
Kevin Windrem avatar image
Kevin Windrem answered ·

I can't find the article but mvader was involved a file system full issue a while back and discovered that node-red had filled up the /data partition with it's log files.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

mvader (Victron Energy) avatar image
mvader (Victron Energy) answered ·

Hi, best check chapter 8 of the venus os large manual, here: https://www.victronenergy.com/live/venus-os:large#troubleshooting.

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

jonoci avatar image jonoci commented ·
Thank you Kevin and mvader,

The data partition is as I said , full of many node-contrib nodes, many of which I did not install, but seem to get installed when you install certain related nodes, eg, twilio for SMS capability installs the SMS capability node, but also a huge heap of other node item s- Whatsapp, SMS Costings, etc - some 180MB of 'stuff' Also, things like Home Assistant, which I never loaded, never use, etc is in the directory, with many other nodes/functions I never incorporated or use. There are many such items - 100's of MB...

There are not many node-red log files that I have found.

Likewise, npm has in its -cacache directories hundreds of tmp files - a npm -cleanup reduces my 96% used space to 95%.... So that did not help, and there are still hundreds of tmp files there - Who knows what you can delete?

I did read chapter 8 of Venus OS large - many times in the past weeks - it helps, but not really...There is no info anywhere I can find that details the directories and files to the extent that once can safely work out what can be deleted and what not - Can I just delete all 'home-assistant' files etc?

I have in essence given up - I purchased another Cerbo GX - this one happens to have 4GB data user space, the old one has 1GB...

I installed the latest firmware, inc Venus OS Large latest. I then imported all my nodes/flows after exporting them from the old Cerbo, and installed all the nodes indicated as missing.

The flows all work fine, except for the flows related to the interface to the external equipment ( Victron inverters/charges/MPPT/Fronious Solar inverter, etc).

Those flows are present, but the cerbo is not yet connected to the equipment.

I am preparing to swap out the new cerbo with the old. I need to do the swap procedure as per the VRM site requirements to maintain my old data. I have taken screen shots of ALL the Remote Console pages so that I have device list and setup data so I can configure the new cerbo with that data ( wish that config file transfer was done...)

The new cerbo, with ALL flows and nodes installed, and without the equipment connected and no config setup yet, shows data space used = 3% ( of 4GB) - so less than 10% of a 1GB space, and the old Cerbo shows 95%...

This is really a bit of a mess and impossible to fix - hence the new cerbo, which is a very expensive solution! I will get the new one running fully, and then factory reset the old one and repeat the process on the old one again and see what data space is left...This is one of those memory leak type situations - nibbling away at your Bytes, never to be recovered again.

0 Likes 0 ·
jonoci avatar image
jonoci answered ·

I completed this silly exercise...

Installed and fully configured a new Cerbo ( this one has 4GB data space, the old cerbo has 1GB), installed OS-Large, latest FW versions, and all my flows. Installed into the Victron system and re-configured the Cerbo for the system ( Fronius 3 phase, 3x Multiplus-II's, MPPT 450/200 , 12xBYD batteries, Heat pump, etc..)

Once all working I ssh'd into root and found 3% memory used (of 4GB data space)

I then removed the Cerbo, set the old cerbo back to factory defaults ( venus-data-90-reset-all.tgz) and check the memory use before and after the reset - Before = 96%, after = 0%.

I then reloaded latest FW, OS-Large and all my flows, inserted it into the Victron system and reconfigured again...

Once running, ssh showed memory use = 11% ( of 1GB data space)

So, exactly the same as the new cerbo ( 3% of 4GB ~ = to 11% of 1GB). I believe there is a 'memory leak' in OS-Large, made worse with large flows...I will keep an eye on the memory growth ( or shrinkage...)

Images are: Old Cerbo 96% full, New Cerbo 3% full, Old cerbo 11% full.

old-cerbo-df-list-pre-reset.png new-cerbo-df-list.png

old-cerbo-df-list-post-reset-post-flows-post-fulli.png


old-cerbo-df-list-pre-reset.png (16.5 KiB)
new-cerbo-df-list.png (28.1 KiB)
old-cerbo-df-list-post-reset-post-flows-post-fulli.png (18.8 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Related Resources

Additional resources still need to be added for this topic