Kener v4 feedback

[rajnandan1]

Hi all

I have deployed the v4 version here

I would really love your feedback. If you want login credentials please dm me on twitter/x with your email I will send a invite

[coodos]

@rajnandan1 one feature that we really miss in the current version is the ability to have categories which are expanded, at the moment if we use categories then the user has to click them in order to view the stats there, what would be really great is if we can have them accordion style, where they stay expanded similar to what atlassian status page does.

Happy to open a PR if this is something you feel would be valuable to you too :)

[rajnandan1]

Can you share a status page that has this. I want to see the animation

[coodos]

see either of these for expandable sections

https://status.digitalocean.com/
https://status.usg.edu/

just to give you an idea this is one of the status pages we have for one of our projects

https://status.w3ds.metastate.foundation
in here we would just want that the categories can expand and it doesnt open on another page

[rajnandan1]

@coodos what do you think of this
https://kener-v4-alpha-production.up.railway.app/samples
check out the group monitor here

[coodos]

This looks amazing, thank you!

[kdambekalns]

Since 4.0.3 is now marked as the latest release, I tried out upgrading. Adjusted my Docker Compose setup to have Redis and when starting Kener I see:

kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ✅ audit secrets and track compliance: https://dotenvx.com/ops
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: 🗂️ backup and recover secrets: https://dotenvx.com/ops
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ⚙️  enable debug logging with { debug: true }
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ⚙️  suppress all logs with { quiet: true }
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ⚙️  override existing env vars with { override: true }
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ⚙️  suppress all logs with { quiet: true }
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: 👥 sync secrets across teammates & machines: https://dotenvx.com/ops
kener  | [dotenv@17.2.4] injecting env (0) from .env -- tip: ⚙️  write to custom object with { processEnv: myObject }
kener  | Running migrations...
kener  | Error running migrations: Error: The migration directory is corrupt, the following files are missing: 20250111153517_init.js, 20250201152526_add_new_column_to_incidents.js, 20250222153624_add_incident_source_column.js, 20250315091330_add_table_invitations.js, 20250417144954_add_subscription_tables.js
kener  |     at validateMigrationList (/app/node_modules/knex/lib/migrations/migrate/Migrator.js:567:11)
kener  |     at Migrator.latest (/app/node_modules/knex/lib/migrations/migrate/Migrator.js:69:7)
kener  |     at async runMigrations (file:///app/build/main.js:10227:5)
kener  |     at async Server.<anonymous> (file:///app/build/main.js:10243:3)
kener  | Running seed...
kener  | Error running seed: Error: Error while executing "/app/seeds/generate_template.ts" seed: select count(`template_id`) as `CNT` from `general_email_templates` where `template_id` = 'subscription_account_code' limit 1 - no such table: general_email_templates
kener  |     at Seeder._waterfallBatch (/app/node_modules/knex/lib/migrations/seed/Seeder.js:118:23)
kener  | SqliteError: select count(`template_id`) as `CNT` from `general_email_templates` where `template_id` = 'subscription_account_code' limit 1 - no such table: general_email_templates
kener  |     at Database.prepare (/app/node_modules/better-sqlite3/lib/methods/wrappers.js:5:21)
kener  |     at Client_BetterSQLite3._query (/app/node_modules/knex/lib/dialects/better-sqlite3/index.js:35:34)
kener  |     at executeQuery (/app/node_modules/knex/lib/execution/internal/query-executioner.js:37:17)
kener  |     at Client_BetterSQLite3.query (/app/node_modules/knex/lib/client.js:154:12)
kener  |     at Runner.query (/app/node_modules/knex/lib/execution/runner.js:141:36)
kener  |     at ensureConnectionCallback (/app/node_modules/knex/lib/execution/internal/ensure-connection-callback.js:13:17)
kener  |     at Runner.ensureConnection (/app/node_modules/knex/lib/execution/runner.js:318:20)
kener  |     at async Runner.run (/app/node_modules/knex/lib/execution/runner.js:30:19)
kener  |     at async Module.seed (file:///app/seeds/generate_template.ts:10:15)
kener  |     at async Seeder._waterfallBatch (/app/node_modules/knex/lib/migrations/seed/Seeder.js:115:9) {
kener  |   original: SqliteError: select count(`template_id`) as `CNT` from `general_email_templates` where `template_id` = 'subscription_account_code' limit 1 - no such table: general_email_templates
kener  |       at Database.prepare (/app/node_modules/better-sqlite3/lib/methods/wrappers.js:5:21)
kener  |       at Client_BetterSQLite3._query (/app/node_modules/knex/lib/dialects/better-sqlite3/index.js:35:34)
kener  |       at executeQuery (/app/node_modules/knex/lib/execution/internal/query-executioner.js:37:17)
kener  |       at Client_BetterSQLite3.query (/app/node_modules/knex/lib/client.js:154:12)
kener  |       at Runner.query (/app/node_modules/knex/lib/execution/runner.js:141:36)
kener  |       at ensureConnectionCallback (/app/node_modules/knex/lib/execution/internal/ensure-connection-callback.js:13:17)
kener  |       at Runner.ensureConnection (/app/node_modules/knex/lib/execution/runner.js:318:20)
kener  |       at async Runner.run (/app/node_modules/knex/lib/execution/runner.js:30:19)
kener  |       at async Module.seed (file:///app/seeds/generate_template.ts:10:15)
kener  |       at async Seeder._waterfallBatch (/app/node_modules/knex/lib/migrations/seed/Seeder.js:115:9) {
kener  |     code: 'SQLITE_ERROR'
kener  |   }
kener  | }
kener  | Kener is running on port 3000!
kener  | Using option immediately with every does not affect the job's schedule. Job will run immediately anyway.
kener  | Maintenance event scheduler started (runs every hour)
kener  | Daily cleanup scheduler started (runs at 00:00 UTC)
kener  | Error updating maintenance event statuses: SqliteError: select `maintenances_events`.*, `maintenances`.`title`, `maintenances`.`description` from `maintenances_events` inner join `maintenances` on `maintenances_events`.`maintenance_id` = `maintenances`.`id` where `maintenances_events`.`status` = 'SCHEDULED' and `maintenances`.`status` = 'ACTIVE' and `maintenances_events`.`start_date_time` > 1771944720 and `maintenances_events`.`start_date_time` <= 1771948320 - no such table: maintenances_events
kener  |     at Database.prepare (/app/node_modules/better-sqlite3/lib/methods/wrappers.js:5:21)
kener  |     at Client_BetterSQLite3._query (/app/node_modules/knex/lib/dialects/better-sqlite3/index.js:35:34)
kener  |     at executeQuery (/app/node_modules/knex/lib/execution/internal/query-executioner.js:37:17)
kener  |     at Client_BetterSQLite3.query (/app/node_modules/knex/lib/client.js:154:12)
kener  |     at Runner.query (/app/node_modules/knex/lib/execution/runner.js:141:36)
kener  |     at ensureConnectionCallback (/app/node_modules/knex/lib/execution/internal/ensure-connection-callback.js:13:17)
kener  |     at Runner.ensureConnection (/app/node_modules/knex/lib/execution/runner.js:318:20)
kener  |     at process.processTicksAndRejections (node:internal/process/task_queues:103:5)
kener  |     at async Runner.run (/app/node_modules/knex/lib/execution/runner.js:30:19)
kener  |     at async MaintenancesRepository.getScheduledEventsStartingSoon (file:///app/build/main.js:5958:12) {
kener  |   code: 'SQLITE_ERROR'
kener  | }
kener  |   _  __                            _  _    ___   _____
kener  |  | |/ /___ _ __   ___ _ __  __   _| || |  / _ \ |___ /
kener  |  | ' // _ \ '_ \ / _ \ '__| \ \ / / || |_| | | |  |_ \
kener  |  | . \  __/ | | |  __/ |     \ V /|__   _| |_| | ___) |
kener  |  |_|\_\___|_| |_|\___|_|      \_/    |_|(_)___(_)____/
kener  |
kener  | Kener version 4.0.3 is running!
kener  |
kener  | [500] GET /
kener  | SqliteError: select * from `pages` where `page_path` = '' limit 1 - no such table: pages
kener  |     at Database.prepare (/app/node_modules/better-sqlite3/lib/methods/wrappers.js:5:21)
kener  |     at Client_BetterSQLite3._query (/app/node_modules/knex/lib/dialects/better-sqlite3/index.js:35:34)
kener  |     at executeQuery (/app/node_modules/knex/lib/execution/internal/query-executioner.js:37:17)
kener  |     at Client_BetterSQLite3.query (/app/node_modules/knex/lib/client.js:154:12)
kener  |     at Runner.query (/app/node_modules/knex/lib/execution/runner.js:141:36)
kener  |     at ensureConnectionCallback (/app/node_modules/knex/lib/execution/internal/ensure-connection-callback.js:13:17)
kener  |     at Runner.ensureConnection (/app/node_modules/knex/lib/execution/runner.js:318:20)
kener  |     at process.processTicksAndRejections (node:internal/process/task_queues:103:5)
kener  |     at async Runner.run (/app/node_modules/knex/lib/execution/runner.js:30:19)
kener  |     at async PagesRepository.getPageByPath (file:///app/build/server/chunks/db-Bx3hVLBW.js:1250:12)

It seems my expectation of an automatic migration is supposed to be fulfilled, but fails:

kener  | Running migrations...
kener  | Error running migrations: Error: The migration directory is corrupt, the following files are missing: 20250111153517_init.js, 20250201152526_add_new_column_to_incidents.js, 20250222153624_add_incident_source_column.js, 20250315091330_add_table_invitations.js, 20250417144954_add_subscription_tables.js
kener  |     at validateMigrationList (/app/node_modules/knex/lib/migrations/migrate/Migrator.js:567:11)
kener  |     at Migrator.latest (/app/node_modules/knex/lib/migrations/migrate/Migrator.js:69:7)
kener  |     at async runMigrations (file:///app/build/main.js:10227:5)
kener  |     at async Server.<anonymous> (file:///app/build/main.js:10243:3)

and the following errors are results of the migration failing. The missing files are part of the Git repo, though. 🤔

I made sure to have pulled the latest rajnandan1/kener:latest to no avail.

[rajnandan1]

let me fix that

[rajnandan1]

it is getting built 4.0.5 please clear cache and pull the latest image after 20-30 mins

[kdambekalns]

Yep, migrations now run fine, startup succeeds (using 4.0.8 now). Thanks! 👍

But I see no monitors, just "No Status Available", even though I started this on a copy of the current v3 database… Going to the management and assigning the monitors to the "Home" page (and a new page, reflecting the two categories we have currently) fixed that, too. The last issue was that uploads (site logo, favicon) were not showing up, until I re-uploaded them.

So these things (image re-upload, pages setup) should be made clearer in the upgrading instructions (which are in the changelog, but not at https://kener.ing/docs/migration/v3-to-v4 as linked to from the docs), I did not read

In v4, monitor grouping/display should be managed through Pages.

(emphasis by me) as "must be assigned before they become visible". I'd happily do this, unless you plan to add a migration so that both cases get fixed automagically (should be easy for the images, for the monitors category-to-page it is not as obvious).

[kdambekalns]

Uh-oh… obviously I overlooked

If you used v3 categories, migrate display organization to v4 pages and verify monitor order per page.

[kdambekalns]

More issues…

• We have two font files in /uploads/… currently, and reference them in "Custom CSS" as https://…/uploads/font.woff2. That no longer works, it results in a 404. How to solve this in v4? The link to the related guide has been updated but shows a 404.
• Subscriptions did not survive the upgrade (migration), subscription is turned off and all subscribers are lost.
• Migrated monitors show The "msecs" argument must be of type number. Received type string ('30000') under "Monitoring Data" until the timeout was changed and saved.

Should I create issues for those findings, instead of (mis-)using this discussion?

[fixundfertig123]

Hi, @rajnandan1

thank you for your work. Starting to be able to migrate from kener v3 to kener v4 (more on that later), I think I have a understanding problem concerning groups and pages. Please let me explain it in an simplified example:

Goal:

• I have 5 "categories" and each category has 5 monitors (categories is my abstract term, monitor in the terminology of kener)
• I want to show the status of these categories on the home "page" (categories is abstract term, "page" in the terminology of kener)

To my understanding, in kener v4 I have to setup a kener group-monitor for each of my "categories" each containing 5 monitors. In generally each monitor is weighted 0.2 to sum up to 1. Of course I could change it, but lets assume the 5 monitors are equally important.

Now my problem:
If one monitor fails the status of the group-monitor still says "up". And according to the documentation https://kener.ing/docs/v4/monitors/group#status-scoring-model that expected since the calculation would be 0.2 * 0 + 0.2 * 0 + 0.2 * 0 + 0.2 * 0 + 0.2 * 2 = 0.4. And since 0.4 is below 1 the group monitor is <1 and threreby "up".

Question:

• One out of five servers in a group is down and the group is still saying group is up running? That cannot be expected?!?
• Or is my approach wrong using group monitors for this?

Possible solution approach:
To my understanding in server monitoring maintenance flag should not be part of the server status, since a can server flip between up,down or degraded, but that's irrelevant if the server it is in maintenance mode. Mathematically speaking: maintenance mode is orthogonal the status mode. Maintenance is an additional boolean flag. Thereby my suggestion would be to norm the status between 0 and 1. Zero means down and One means up. Anything in between in degraded.

[fixundfertig123]

And I do not really understand your weight approach. To my understanding your current calculation would always overweigth the "up" vs the down if you have more than 2 monitors in a group. Using 0 as a weight for "up" status you have not normalized the values.

As said using 0 and 1 for up and down (and not considering mainteance) and >0 and <1 for degrading would make more sense. Since a group is down only if all servers are down. Otherwise a group is only degraded, since some servers are still running.

[rajnandan1]

changed in v4.0.15

https://kener.ing/docs/v4/changelogs/v4.0.14

[fixundfertig123]

@rajnandan1 That's great news. Seems to work perfectly. I keep you updated. Thank you very much! Just bought some coffees ;-)

[rajnandan1]

@fixundfertig123 thank you so much. means a lot.

[fixundfertig123]

@rajnandan1 I have to say thanks. You are helping me. So both of us should benefit :-)

[fixundfertig123]

Hi @rajnandan1,
I have tested a bit. And I observed an issue and having a thought it might be not trivial.

My Situation:

• I have X monitors of type manual which are by default state "up". I have an internal system that updates these monitors manually by calling by API.
• I have a group monitor, that includes these X monitors, equally weighted.

My observation:

• Each monitor gets a udpated every minute by kener and set to default "up". (works as expected)
• Now my system may updates a monitor to "down" given the proper timestamp. But there can be a delay of lets say 10 seconds between the API update and the "native" kener default update.

Result:
Monitor is down, but group is still up, see picture and yellow marks

Question:
Am I doing something wrong? I played with with execution delay, but it did not work and and the parameter is not so well explained.

[rajnandan1]

Let me fix that, I am sorry but the execution delay is not working as expected. I will fix that.
In your case it will be execution delay 11 seconds

[fixundfertig123]

Thanks for the quick reply. That's what I thought about the execution delay. I experimented a bit but unfortunately not working as expected. If I can help somehow, just let me know.

I hope to dive more into the code sooner to help a little bit more proactive, but unfortunately at the moment I am drowning in work...

[rajnandan1]

I have raised a PR. Execution tumour was working fine but api updates were not getting propagated to the group. Will deploy soon.

[rajnandan1]

deployed: https://kener.ing/docs/v4/changelogs/v4.0.17#date-time-formats

[fixundfertig123]

@rajnandan1 Thank you very much! I really appreciate it. Looks good so far.

[fixundfertig123]

Hi, I just looked into the code a bit and to my understanding "manual" monitors should also have a execution delay implemented (between 1000 and e.g 30000ms but of course less than 60000ms) for the same reason as above. Otherwise you would have a permanent flipping between "up" state - triggered by kener cron - and "down" state - triggered by API call if the time interval for API updates is set to minutes.

[k4nib4l3]

First of all, thank you very much for the great work, I really appreciate it.
I am currently running Kener version 3.2.19 via Docker. I have tried several times to upgrade to v4, but unfortunately I run into the same issue every time: the login page and the /manage subpage return a 404 error, which leaves me unable to use the system.
The main status page works without any problems, although my monitors, users, etc. from version 3.2.19 are not carried over.
Do you have any idea why the management subpages stop working after the upgrade?
Thank you so much!

[fixundfertig123]

Hi, does it just give a 404 or some more details? Does the kener log say something?

[k4nib4l3]

I just figured it out finally. The manage url has been changed with v4. It is now "/account/signin" instead of /manage.

[fixundfertig123]

Ah. Sorry. Totally forgot about that. Run into the same problem. Sorry for not saving you some time.

[rajnandan1]

sorry i think i did not add it in changelogs. i will add it now