-
Notifications
You must be signed in to change notification settings - Fork 1.7k
Do not use the home directory as the working dir in docker #9834
Conversation
It looks like @fubhy hasn't signed our Contributor License Agreement, yet.
You can read and sign our full Contributor License Agreement at the following URL: https://cla.parity.io Once you've signed, please reply to this thread with Many thanks, Parity Technologies CLA Bot |
[clabot:check] |
It looks like @fubhy signed our Contributor License Agreement. 👍 Many thanks, Parity Technologies CLA Bot |
To be honest, I can't think of a scenario where one would want to run the docker image without mounting the base path and passing a custom path to --base-path. On the other hand, it does not hurt to create the home directory as long as we don't move any binaries or such there. |
For my curiosity, what was the problem in k8s with having the binary under |
Because by default, kubernetes runs under a random UID and GID as an additional security precaution. K8s suggests to not rely.on the USER directive at all. It's possible to configure your way around it of course but that's a bit of a uneccesary hassle |
This broke our docker builds |
* Do not create a home directory. * Re-add -m flag
* version: bump stable to 2.1.7 * Adjust requests costs for light client (#9925) * PIP Table Cost relative to average peers instead of max peers * Add tracing in PIP new_cost_table * Update stat peer_count * Use number of leeching peers for Light serve costs * Fix test::light_params_load_share_depends_on_max_peers (wrong type) * Remove (now) useless test * Remove `load_share` from LightParams.Config Prevent div. by 0 * Add LEECHER_COUNT_FACTOR * PR Grumble: u64 to u32 for f64 casting * Prevent u32 overflow for avg_peer_count * Add tests for LightSync::Statistics * Fix empty steps (#9939) * Don't send empty step twice or empty step then block. * Perform basic validation of locally sealed blocks. * Don't include empty step twice. * prevent silent errors in daemon mode, closes #9367 (#9946) * Fix light client informant while syncing (#9932) * Add `is_idle` to LightSync to check importing status * Use SyncStateWrapper to make sure is_idle gets updates * Update is_major_import to use verified queue size as well * Add comment for `is_idle` * Add Debug to `SyncStateWrapper` * `fn get` -> `fn into_inner` * ci: rearrange pipeline by logic (#9970) * ci: rearrange pipeline by logic * ci: rename docs script * Add readiness check for docker container (#9804) * Update Dockerfile Since parity is built for "mission critical use", I thought other operators may see the need for this. Adding the `curl` and `jq` commands allows for an extremely simple health check to be usable in container orchestrators. For example. Here is a health check for a parity docker container running in Kubernetes. This can be setup as a readiness Probe that would prevent clustered nodes that aren't ready from serving traffic. ```bash #!/bin/bash ETH_SYNCING=$(curl -X POST --data '{"jsonrpc":"2.0","method":"eth_syncing","params":[],"id":1}' http://localhost:8545 -H 'Content-Type: application/json') RESULT=$(echo "$ETH_SYNCING | jq -r .result) if [ "$RESULT" == "false" ]; then echo "Parity is ready to start accepting traffic" exit 0 else echo "Parity is still syncing the blockchain" exit 1 fi ``` * add sync check script * Fix docker script (#9854) * Dockerfile: change source path of the newly added check_sync.sh (#9869) * Do not use the home directory as the working dir in docker (#9834) * Do not create a home directory. * Re-add -m flag * fix docker build (#9971) * bump smallvec to 0.6 in ethcore-light, ethstore and whisper (#9588) * bump smallvec to 0.6 in ethcore-light, ethstore and whisper * bump transaction-pool * Fix test. * patch cargo to use tokio-proto from git repo this makes sure we no longer depend on smallvec 0.2.1 which is affected by servo/rust-smallvec#96 * use patched version of untrusted 0.5.1 * ci: allow audit to fail
Moving the binary and the entry point to the home directory makes using the docker image in Kubernetes really hard. At the same time, there is no good reason to do so as far as I can tell (all that we want is to use a non-root user which we can absolutely do even without the home directory).