pfs | pCloud filesystem client | File Utils library

The problem was caused by jupyter/tensorflow being loaded in the background. The following solved the issue:

The issue is your map() call. It will return an Object (your org) if it has at least one client that passes the filter(), but implicity returns undefined if filtered.length is not greater than zero.

In JS every function has a return -- either declared with return, or implicit. The default implicit value is undefined which is where those array entries are coming from. You could switch map() to filter() and it should work (since it will omit the falsy undefined values from the array).

Yes, there is a way to remove this. The answer is quite straightforward.

Just use the Custom GelfLayout. You can use the configuration from above but remove the includeStackTrace, includeThreadContext and customFields line:

Rather than trying to ignore the namespaces, if there is only a default namespace, you can remove the xmlns attribute from the document root:

Is TLS (1.2) enforced for Cognito:

Unfortunately, no. For most API calls you could use API gateway as a layer in between and enforce TLS1.2 there. However, for the AUTHORIZATION and TOKEN endpoints this does not work. Alternatively, you can enforce it for all endpoints by deploying cognito to the US regions and using the FIPS endpoints of cognito there. See https://docs.aws.amazon.com/general/latest/gr/cognito_identity.html for the available fips endpoints.

with encryption in transit being available would security best practices still dictate having Cognito available through VPC endpoint so that the token does not need to travel on internet:

That all depends on your risk apetite. For most organizations, encryption will be sufficient. However if your risk apetite is lower, you'll want extra mitigations, following the "defense in depth" best practice. Private connectivity is just one option if many you could add. Monitoring on suapicious user activity and adaptable MFA are other methods for example.

if a client application is coming from on-prem environment and accessing the api then if we make it go through a proxy in AWS then will the traffic remain in AWS backbone network or still flow through internet?

As the api endpoints are public, they will go over the internet (public ip address, so it's routed to a NAT gateway or internet gateway).

Adding public key to instance metadata by the username core seems to solve the issue. Any other username is rejected.

1. Generate key pair: ssh-keygen -t ed25519
2. Copy public key to instance metadata - link.
3. Restart the instance.
4. Connect using new key: ssh -i core@

More information about this issue can also be found here.

First - depending on the image you're running your machine from - it has to have tmux installed. If it's a new machine with Debian 10 you need to put sudo apt install tmux -y at the start of your startup script to install it.

To check if the script ran at the start you can add the touch /tmp/testfile1.txt at the end and when the VM has booted up check if the file exists. That's the easies (and not so reliable way to tell if the script ran).

I'm not familiar with tmux but I've found out that the server service will exit of there are no sessions created, it looks to me like the server exitx before the new sessions are established. You can try using sleep 1 suggested here to solve your issue.

I tried running your script as is but had the same results as you, but I did the debugging I mentioned and everything worked;

I added some "debugging" lines to the script and ran it:

The values in your dictionary are lists with strings in them. So dict.values() returns a list of lists. When you use .join, it is converting that list into a str instead of the elements of that list:

Rules for memory allocation.

• Do not allocate so much RAM that swapping will occur. Swapping is terrible for MySQL/MariaDB performance.
• Do adjust innodb_buffer_pool_size such that most of RAM is in use during normal time and even for spikes in activity. (I often say "set it to 70% of available RAM", but you are asking for more details.)
• Do not bother changing other settings; they add to the complexity of "getting it right".

There are 3 situations (based on innodb_buffer_pool_size and dataset size):

• Tiny dataset -- buffer_pool is bigger than necessary --> wasting some of RAM, but so what; it is not useful for anything else. And it give you some room for growth.
• Medium-sized dataset -- Most activity is done in RAM; the system will run nicely.
• Big dataset -- The system may be I/O-bound. Adding RAM is a costly and brute force solution. However, some software techniques (eg, better indexes) may help, such as this for WordPress and WooCommerce.