cardano-sl crashes during sync

[ghost]

Hi,

I have built a container as of this guide in a VM (using cardano-sl-explorer-static. in the nix-build command). I use version 3.0.2. I start it with the cardano-explorer command.
It starts fine and runs for about 30 minutes but then it crashes and the container stops. Sometimes it says "Failed to add headers from ... to block retrieval queue: queue is full" in the log but not always.
What I've tried so far:

• restart container
• add more RAM, reboot, restart container
• add more CPU, reboot, restart container
  None of them works, I cleared the database before each restart. Please help me, logs attached
  Cardano_sync_error_log.txt
  Cardano_sync_error_log_2.txt
  Many thanks.

[Victoryfan]

I also meet this problem. Do you resolve it? Anyone can help me?

[erikd]

How much RAM does the VM have? The IOHK devops team runs one on a 64Gig VM and if still runs out of memory, crashes and is restarted automatically.

The existing explorer is of very poor quality and we are working on providing a new one.

[Victoryfan]

How much RAM does the VM have? The IOHK devops team runs one on a 64Gig VM and if still runs out of memory, crashes and is restarted automatically.

The existing explorer is of very poor quality and we are working on providing a new one.

But during syncing data, it has bugs as picture

[erikd]

The "attempted to subtract" problem is know and has been looked at. We could not get to the bottom of it, and decided that since the explorer is going to be replaced anyway, it was not worth putting any more dev effort into this one.

Address balances pulled from the explorer can be incorrect. Data about block and transactions is thought to be 100% correct.