So far, only calls that go to the Java backend, had the
`X-Content-Type-Options` header in the responses. This PR adds them to
all responses by
1. adding it to NGINX configuration.
2. removing it from Spring security's configuration, so we don't end up
with _two_ `X-Content-Type-Options` headers in the response.
---------
Co-authored-by: Nidhi <nidhi@appsmith.com>
When Appsmith is deployed with HTTPS, all HTTP requests are redirected
to HTTP. But this redirection response, is leaking the NGINX version in
the `Server` header. This PR fixes that.
The health endpoint in the Docker container's healthcheck command is
currently pointing to `/users/me`. But we have a designated endpoint for
this, at `/health`. This PR switches to using this endpoint instead.
This will allow us to
1. Bake different CS URLs for release and master builds.
2. Be resilient to the CS URL being set to empty string, as opposed to
not being set at all.
When `APPSMITH_CLOUD_SERVICES_BASE_URL` is an empty string, Spring's
interpolation in `application.properties` doesn't take the default
value. This PR sets up the `entrypoint.sh` script so that it'll `unset`
this env variable, when it's empty.
1. Changing the Maps API Key doesn't need restart anymore.
2. The `isRestartRequired` field in the response of updating env
settings, was being ignored. The client owns the decision of when to
restart (which is correct), so removed this from the server.
3. Write Maps API Key to the database, in the tenant configuration.
4. The Settings page for Maps Ke gets the current value from
`/tenant/current` response, and not `/admin/env`.
5. Removed `APPSMITH_GOOGLE_MAPS_API_KEY` from `/admin/env` response.
6. Tests.
DO NOT MERGE. Please only review/approve. This is expected to break EE
once it goes there, which I intend to solve alongside merging this.
Changing the Maps API Key will update it both in the tenant config in
the database, as well as in the `docker.env` file. This is predominantly
for backwards compatibility, and phased rollout. As part of a separate
PR, we'll have a migration that proactively copies the env variable
value to the database, and comment out the value in the `docker.env`
file. Then we can stop updating the `docker.env` file as well.
## New
## Old
---------
Co-authored-by: Ankita Kinger <ankita@appsmith.com>
We've stopped [building the slim
images](https://github.com/appsmithorg/appsmith/pull/25219) now and so
the instructions to run server locally, which rely on the slim server
image, need to be updated to use the newer `appsmith-ce` image.
This PR updates the instructions to use this.
## Description
- The default request timeout on nginx is set to 60s. This feat allows
users to configure custom nginx timeout value specified via the
Environment variable `APPSMITH_SERVER_TIMEOUT`
- Solves https://github.com/appsmithorg/appsmith/issues/14535
## Testing
#### How Has This Been Tested?
- [x] Manual
- [ ] Jest
- [ ] Cypress
---------
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
This will complete:
[23522](https://github.com/appsmithorg/appsmith/issues/23522)
Currently, when we deploy appsmith on cloudrun we see this error. This
commit will fix this issue
```
2023-06-14 12:32:10.513 IST
Mounting File Sytem
2023-06-14 12:32:10.521 IST
mount.nfs: mount point /appsmith-stacks does not exist
2023-06-14 12:32:10.533 IST
Container called exit(32).
```
Fixes: [24013](https://github.com/appsmithorg/appsmith/issues/24013)
The issue started with the introduction of adding of startup-events.
Because of the ipv6 header
```
server {
listen 80 default_server;
listen [::]:80 default_server;
}
```
present in the
`/etc/nginx/sites-enabled/default` and calling nginx
[here](https://github.com/appsmithorg/appsmith/blob/release/deploy/docker/entrypoint.sh#L392).
With this change we are removing ```listen [::]:80 default_server;```
from the file before starting nginx.
---------
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
- Removed event listener python script, removed supervisor conf for the
same.
- Added new python script to handle the starting page init.
- Python script will be called via a shell script with a failsafe
removal of the loading page.
- The above shell script will be run as a child of the run java script
Tests:
The child script exits after execution, and does not create zombie
process even when the run java script is called multiple times with
restarts.
Verified exceptions and failsafe manually by introducing syntax errors
and uninstalling requests module.
---------
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
## Description
> Add a TL;DR when description is extra long (helps content team)
>
> Please include a summary of the changes and which issue has been
fixed. Please also include relevant motivation
> and context. List any dependencies that are required for this change
>
> Links to Notion, Figma or any other documents that might be relevant
to the PR
>
>
#### PR fixes following issue(s)
Fixes # (issue number)
> if no issue exists, please create an issue and ask the maintainers
about this first
>
>
#### Media
> A video or a GIF is preferred. when using Loom, don’t embed because it
looks like it’s a GIF. instead, just link to the video
>
>
#### Type of change
> Please delete options that are not relevant.
- Bug fix (non-breaking change which fixes an issue)
- New feature (non-breaking change which adds functionality)
- Breaking change (fix or feature that would cause existing
functionality to not work as expected)
- Chore (housekeeping or task changes that don't impact user perception)
- This change requires a documentation update
>
>
>
## Testing
>
#### How Has This Been Tested?
> Please describe the tests that you ran to verify your changes. Also
list any relevant details for your test configuration.
> Delete anything that is not relevant
- [ ] Manual
- [ ] Jest
- [ ] Cypress
>
>
#### Test Plan
> Add Testsmith test cases links that relate to this PR
>
>
#### Issues raised during DP testing
> Link issues raised during DP testing for better visiblity and tracking
(copy link from comments dropped on this PR)
>
>
>
## Checklist:
#### Dev activity
- [ ] My code follows the style guidelines of this project
- [ ] I have performed a self-review of my own code
- [ ] I have commented my code, particularly in hard-to-understand areas
- [ ] I have made corresponding changes to the documentation
- [ ] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] New and existing unit tests pass locally with my changes
- [ ] PR is being merged under a feature flag
#### QA activity:
- [ ] [Speedbreak
features](https://github.com/appsmithorg/TestSmith/wiki/Test-plan-implementation#speedbreaker-features-to-consider-for-every-change)
have been covered
- [ ] Test plan covers all impacted features and [areas of
interest](https://github.com/appsmithorg/TestSmith/wiki/Guidelines-for-test-plans/_edit#areas-of-interest)
- [ ] Test plan has been peer reviewed by project stakeholders and other
QA members
- [ ] Manually tested functionality on DP
- [ ] We had an implementation alignment call with stakeholders post QA
Round 2
- [ ] Cypress test cases have been added and approved by SDET/manual QA
- [ ] Added `Test Plan Approved` label after Cypress tests were reviewed
- [ ] Added `Test Plan Approved` label after JUnit tests were reviewed
---------
Co-authored-by: Trisha Anand <trisha@appsmith.com>
Co-authored-by: Nidhi <nidhi@appsmith.com>
Co-authored-by: Arpit Mohan <mohanarpit@users.noreply.github.com>
Co-authored-by: Sumit Kumar <sumit@appsmith.com>
Co-authored-by: ChandanBalajiBP <104058110+ChandanBalajiBP@users.noreply.github.com>
Co-authored-by: Aishwarya-U-R <91450662+Aishwarya-U-R@users.noreply.github.com>
Co-authored-by: balajisoundar <balaji@appsmith.com>
Co-authored-by: Aishwarya UR <aishwarya@appsmith.com>
Co-authored-by: sidhantgoel <sidhant@hexcod.in>
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
Co-authored-by: Satish Gandham <satish.iitg@gmail.com>
Co-authored-by: Nilesh Sarupriya <nilesh@appsmith.com>
Co-authored-by: Nilesh Sarupriya <20905988+nsarupr@users.noreply.github.com>
Co-authored-by: nilansh <nilansh@appsmith.com>
Co-authored-by: Ayangade Adeoluwa <37867493+Irongade@users.noreply.github.com>
Co-authored-by: akash-codemonk <67054171+akash-codemonk@users.noreply.github.com>
Co-authored-by: Vijetha-Kaja <119562824+Vijetha-Kaja@users.noreply.github.com>
Co-authored-by: Manish Kumar <107841575+sondermanish@users.noreply.github.com>
Co-authored-by: Sumesh Pradhan <sumesh@appsmith.com>
Co-authored-by: Rishabh Rathod <rishabh.rathod@appsmith.com>
Co-authored-by: Shrikant Sharat Kandula <shrikantsharat.k@gmail.com>
Co-authored-by: Tanvi Bhakta <tanvi@appsmith.com>
Co-authored-by: Somangshu Goswami <somangshu.goswami1508@gmail.com>
Co-authored-by: Tolulope Adetula <31691737+Tooluloope@users.noreply.github.com>
Co-authored-by: Anagh Hegde <anagh@appsmith.com>
Co-authored-by: Saroj <43822041+sarojsarab@users.noreply.github.com>
Co-authored-by: Nayan <nayan@appsmith.com>
Co-authored-by: Arsalan <arsalanyaldram0211@outlook.com>
Co-authored-by: ashit-rath <ashit@appsmith.com>
Co-authored-by: Rimil Dey <rimil@appsmith.com>
Co-authored-by: Druthi Polisetty <druthi@appsmith.com>
Co-authored-by: Sangeeth Sivan <74818788+berzerkeer@users.noreply.github.com>
Co-authored-by: subratadeypappu <subrata@appsmith.com>
Co-authored-by: Ayush Pahwa <ayush@appsmith.com>
Co-authored-by: Rajat Agrawal <rajat@appsmith.com>
Co-authored-by: Pawan Kumar <pawan.stardust@gmail.com>
Co-authored-by: Abhijeet <41686026+abhvsn@users.noreply.github.com>
Co-authored-by: Aman Agarwal <aman@appsmith.com>
Co-authored-by: Parthvi Goswami <parthvigoswami@Parthvis-MacBook-Pro.local>
Co-authored-by: Anand Srinivasan <66776129+eco-monk@users.noreply.github.com>
Co-authored-by: Anand Srinivasan <anand.a.srinivasan@outlook.com>
Co-authored-by: Satish Gandham <hello@satishgandham.com>
Co-authored-by: Goutham Pratapa <goutham@appsmith.com>
Co-authored-by: manish kumar <manish@appsmith.com>
Co-authored-by: Dipyaman Biswas <dipyaman@appsmith.com>
Co-authored-by: Ayush Pahwa <ayushpahwa96@gmail.com>
Co-authored-by: Rohit Agarwal <rohit_agarwal@live.in>
Co-authored-by: Tanvi Bhakta <tanvibhakta@gmail.com>
Co-authored-by: Ankit Srivastava <67647761+ankitsrivas14@users.noreply.github.com>
Co-authored-by: tkAppsmith <131347120+tkAppsmith@users.noreply.github.com>
We replace env variables in `index.html`, `view.html` and `edit.html`,
just before start NGINX. But the `.gz` versions of these files don't
have these changes applied to them. This PR gzips those HTML files
again, after the substitutions are applied.
From
https://github.com/appsmithorg/appsmith/pull/23539#issuecomment-1554509537.
- Added cron task to cleanup log files older than 7 days to run weekly.
- Updated supervisor log rotation config to store only 2 backups for all
processes.
## Description
We are solving 2 problems with this PR:
1. For `Installation Setup Complete` event email domain hash was
incorrect and earlier we were hashing the instanceId which means if user
with same email tries to run 2 different instances hash was different.
2. `IP` is a reserved keyword for tracking events in Mixpanel though
this is allowed in Segment. Instead of showing the ip as is, Mixpanel
provides derived property. As we want derived props alongwith the ip
address we are sharing the IP address in separate keys.
Ref:
https://help.mixpanel.com/hc/en-us/articles/360001355266-Event-Properties
#### PR fixes following issue(s)
Fixes https://github.com/appsmithorg/cloud-services/issues/675
#### Type of change
- Bug fix (non-breaking change which fixes an issue)
#### How Has This Been Tested?
- [x] Manual
## Checklist:
#### Dev activity
- [x] My code follows the style guidelines of this project
- [x] I have performed a self-review of my own code
- [x] I have commented my code, particularly in hard-to-understand areas
- [ ] I have made corresponding changes to the documentation
- [x] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] New and existing unit tests pass locally with my changes
- [ ] PR is being merged under a feature flag
#### QA activity:
- [ ] [Speedbreak
features](https://github.com/appsmithorg/TestSmith/wiki/Test-plan-implementation#speedbreaker-features-to-consider-for-every-change)
have been covered
- [ ] Test plan covers all impacted features and [areas of
interest](https://github.com/appsmithorg/TestSmith/wiki/Guidelines-for-test-plans/_edit#areas-of-interest)
- [ ] Test plan has been peer reviewed by project stakeholders and other
QA members
- [ ] Manually tested functionality on DP
- [ ] We had an implementation alignment call with stakeholders post QA
Round 2
- [ ] Cypress test cases have been added and approved by SDET/manual QA
- [ ] Added `Test Plan Approved` label after Cypress tests were reviewed
- [ ] Added `Test Plan Approved` label after JUnit tests were reviewed
Because tracing is turned on in the `entrypoint.sh`, everything we
source also log all names and values, including the `docker.env` file.
This means all env values are printed to the logs.
This PR disables the tracing to fix this. It was enabled originally for
improved logging information for some of the issues we were facing at
the time, but if we still need such logging information, we need to log
those explicitly with `echo`, instead of using `xtrace`.
Event Listener Python script fixes:
- Added exception handling in health check request
- Updated backend health check point to
`http://localhost:8080/api/vi/health`
**Fixes:**
- Overwriting MongoDB env data on appsmithctl restore.
- appsmithctl backup does not include the MongoDB env vars.
**Features:**
- appsmithctl backup now does not stop backend and rts service during
backup operation.
_Note:-
It will be the responsibility of the Admin to make sure that a manual
backup is really a desired snapshot when done via
the appsmithctl backup util command
To ensure this, Admins may require to stop the backend and rts processes
in the container before a manual backup( We will need to document this)
[ In future we can add an appsmithctl cmd to put appsmith to maintenance
mode with a maintenance page ]_
Fixes # [21603](https://github.com/appsmithorg/appsmith/issues/21603)
We're already checking for the AVX instruction, and producing a helpful
message about it. But, we're doing that check _after_ we invoke the
`mongod` process, which fails on such processes, and the error message
is never shown.
This PR changes this, by doing the check before that `mongod` command is
run.
If the replicaSet check fails, we see the following error:
```
readWrite command not found
```
This is because of using the command interpolation backticks in the
double-quoted string.
## Description
This PR includes changes for self-hosted instances to utilize an
embedded postgres database for the Template's mockdb.
After this change, by default Appsmith will always run the embedded
Postgres, and connect to it instead of the mockdb cloud instance.
**Solves:**
Issue[#20107](https://github.com/appsmithorg/appsmith/issues/20107)
**Changes:**
- [x] Install Postgresql in the Docker container using Dockerfile :-
_Updated Docker file to install PostgreSQL v13_
- [x] Add provision to stop/disable Postgresql from booting based on an
environment variable :- _To disable embedded postgres, set the **ENV**
var `APPSMITH_ENABLE_EMBEDDED_DB` to `0`; Configured supervisord to
control the lifecycle of the postgres server ( Supervisor uses user
postgres to start the server since root is not allowed)_
- [x] Seed data into the Postgresql DB during first container startup.
This ensures that we don't have to ship the Docker
container with data & bloat it unnecessarily. :- _Curretly, the SQL dump
is baked into the image._
- [x] Persist the Postgresql data in /appsmith-stacks. This ensures that
the data is persisted across version upgrades. :- _The postgres uses the
file system at` /stacks/data/postgres/main` ; note: The file system
owner is user/group postgres_
- [x] Update the mock db endpoint in the product to use the local DB by
default instead of the hosted DB for self-hosted instances. The cloud
instance should still use the hosted DB :- _The embedded postgres
authentication is set to the type `trust`, allowing the existing
template mockdb meta data to be used without any update to the existing
templates mockdb endpoint and credentials_
**Note: There is an additional table `mydb` added to the mockdb for
quickly verifying the db source is the embedded postgres, since the
cloud mockdb does not have that table.**
---------
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
When services within the fat container try to communicate with each
other, like backend to RTS etc., if they use the loopback address of
IPv4, `127.0.0.1`, it works. But if they use the loopback address of
IPv6, `::1`, it fails because the NGINX inside the fat container isn't
set to bind to IPv6.
This PR fixes this.
In EE, we attempt to make connections to Keycloak without setting the
hostname on the `WebClient`. This picks up the hostname of `::1` on
systems with IPv6 as default, and so the communication between backend
and Keycloak fails.
This is affecting users on ECS Fargate, for example.
Appsmith supports running with a HTTP proxy, that can be configured with
`HTTP_PROXY` or `HTTPS_PROXY` env variables. Like this:
```sh
HTTP_PROXY=http://myproxy:8080
HTTPS_PROXY=http://myproxy:8443
```
However, this proxy support didn't support authentication. This PR
implements that. Now, proxy with authentication can be configured like
this:
```sh
HTTP_PROXY=http://user:password@myproxy:8080
HTTPS_PROXY=http://user:password@myproxy:8080
```
This is not syntax or standards invented by Appsmith. This is the
standard way proxy is usually configured.
Fixes#16330🍰
## Description
We're increasing the default limit of request payload on cloud so that
100 MB files that are base 64 encoded can also be uploaded via Appsmith.
Fixes#20424
## Type of change
- Bug fix (non-breaking change which fixes an issue)
## How Has This Been Tested?
- Manual
### Test Plan
> Add Testsmith test cases links that relate to this PR
### Issues raised during DP testing
> Link issues raised during DP testing for better visiblity and tracking
(copy link from comments dropped on this PR)
## Checklist:
### Dev activity
- [ ] My code follows the style guidelines of this project
- [ ] I have performed a self-review of my own code
- [ ] I have commented my code, particularly in hard-to-understand areas
- [ ] I have made corresponding changes to the documentation
- [ ] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] New and existing unit tests pass locally with my changes
- [ ] PR is being merged under a feature flag
### QA activity:
- [ ] Test plan has been approved by relevant developers
- [ ] Test plan has been peer reviewed by QA
- [ ] Cypress test cases have been added and approved by either SDET or
manual QA
- [ ] Organized project review call with relevant stakeholders after
Round 1/2 of QA
- [ ] Added Test Plan Approved label after reveiwing all Cypress test
When running on CPUs without an AVX instruction, the embedded MongoDB
cannot run. It fails to run with a cryptic error message, and they are
forced to reach out to support for help.
This PR adds a check, and a warning message, if trying to use embedded
MongoDB on a CPU without the AVX instruction.
Issue: Nginx and RTS used the same env PORT for binding it's service,
while the backend server had the rts port hardcoded on its rts uri.
- Renamed env PORT to APPSMITH_RTS_PORT for starting the rts server.
- Updated nginx config templates to use env `APPSMITH_RTS_PORT`
- Added appsmith.rts.port property in server to use env
APPSMITH_RTS_PORT
- Updated CommonConfig.java rtsBaseDomain to use appsmith.rts.port
---------
Co-authored-by: Shrikant Sharat Kandula <shrikant@appsmith.com>
When we have `NO_PROXY=localhost,one.com,two.com,three.com`, then the
current implementation will pass the following to the backend server:
```
localhost|one.com,two.com,three.com
```
This will mean that only `localhost` will bypass the proxy, which is not
what's expected.
This PR fixes this problem, so that the following is sent to the backend
server:
```
localhost|one.com|two.com|three.com
```
This will mean that requests to all four of them, will bypass the proxy,
as expected.
In the `entrypoint.sh` script, we check if the MongoDB in use, has
replicaSet initiated or not. This is usually done with a `rs.initiate()`
on the cluster.
We need the replicaSet to be enabled on MongoDB, since the backend
server relies on MongoDB `changeStream`s, which is a feature, only
available if replicaSet is enabled.
However, to use the `changeStream` APIs, having the `read` or
`readWrite` role on MongoDB is enough. But the check we do in
`entrypoint.sh`, runs `rs.status()` to see if `replicaSet` is initiated.
This `rs.status()` call, unfortunately, requires the `ClusterMonitor`
role, unlike the `changeStream` API.
To tackle this, we created the `appsmithctl check_replica_set` command.
This command would attempt to use the `changeStream` API, and report
success or failure. But this failed on certain configurations, where
MongoDB was running as a single-node-cluster, on localhost, or a
local/internal network. This was an edge case.
That edge case is solved by this PR. With this, we can use `appsmithctl
check-replica-set` in the `entrypoint.sh` again.
We're `watchtower:latest-dev` image for the `--schedule` support. This
was only available in the `latest-dev` version of Watchtower.
Now, this feature is available in the `latest` version of Watchtower, so
we should be fine switching back to `latest`.
More importantly, the `latest-dev` version of Watchtower, doesn't have
ARM images built. It only has AMD images built today. So this watchtower
service fails to start on ARM architectures. This PR fixes that.
Fixes#19969.
The previous PR that fixed this was failing with the way we were
creating the initial user on MongoDB. This PR addresses that.
Co-authored-by: Aishwarya UR <aishwarya@appsmith.com>
## Description
The Appsmith Docker image was unable to run on CPUs without AVX
instructions because the entrypoint script was using the legacy `mongo`
client. This PR replaces the `mongo` client with the new `mongosh`
client in the entrypoint, which does not require these instructions. Wit
this fix, the Appsmith Docker image is able to run on CPUs without AVX
instructions.
Fixes#19969
## Type of change
- Bug fix (non-breaking change which fixes an issue)
## How Has This Been Tested?
- Manual
## Checklist:
### Dev activity
- [x] My code follows the style guidelines of this project
- [x] I have performed a self-review of my own code
- [ ] I have commented my code, particularly in hard-to-understand areas
- [ ] I have made corresponding changes to the documentation
- [x] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [x] New and existing unit tests pass locally with my changes
- [ ] PR is being merged under a feature flag
### QA activity:
- [ ] Test plan has been approved by relevant developers
- [ ] Test plan has been peer reviewed by QA
- [ ] Cypress test cases have been added and approved by either SDET or
manual QA
- [ ] Organized project review call with relevant stakeholders after
Round 1/2 of QA
- [ ] Added Test Plan Approved label after reveiwing all Cypress test
## Description
- Updated redis.conf to set the Redis rdb directory to
`/appsmith-stacks/data/redis` bind-mount directory, so that the data is
persisted after the container restarts/updates. ( This was causing user
sessions on Appsmith installations using the embedded Redis to reset
after every container restart, since the redis data was not being
persisted)
- Updated redis.conf for the embedded Redis to increase the snapshot
frequency to dump after every 15 sec if at least 1 key changed
Fixes#19768
## Type of change
- Bug fix (non-breaking change which fixes an issue)
## How Has This Been Tested?
- Manual
