tomaszb
@tomaszb
Regular
Regular · Member since Oct 2026
0 followers · 0 following
- Topics created
- 4
- Posts created
- 4
- Likes given
- 8
- Likes received
- 7
- Days visited
- 0
- Solutions
- 0
Days visited count across the whole community, not just this forum.
Most active topics
- Rails 8 solid_queue on Larkhost: one pod or two?
1 replies · 0 votes
- Postgres cluster has no writable primary, all nodes in recovery
1 replies · 0 votes
- How to retrieve boot diagnostics for a destroyed pod?
0 replies · 0 votes
- Does the proxy ever buffer request bodies to disk?
0 replies · 0 votes
Most liked posts
- How to retrieve boot diagnostics for a destroyed pod?
An ephemeral pod failed during boot and was destroyed. lark logs has nothing because it never started the app. Are boot-stage logs kept anywhere?
3 likes
- Postgres cluster has no writable primary, all nodes in recovery
After a region maintenance window, all three nodes of my Postgres cluster report in recovery. Nothing accepts writes. $ lark pg status -a db-prod NODE
2 likes
- HTTP POST to a burrow URL does not wake a paused burrow (GET does)
Interesting — is the body large? Maybe the proxy gives up buffering it while the burrow wakes.
1 likes
- Pods can now be cordoned before a deploy
Does an autostopped pod stay cordoned when it wakes up?
1 likes
Top categories
Top tags
- #logs 1
- #pods 1
- #postgres 1
- #processes 1
- #proxy 1
- Laravel queues stop processing after a deploy
After every deploy, the queue worker pod runs but processes nothing until I restart it by hand.
Likes given · 10/12/2026
- Show & Tell: a tiny job runner on pods that sleep between jobs
Using this for nightly exports now, thanks!
Likes given · 10/12/2026
- Outbound mTLS to a partner API hangs ~70 seconds then fails (iad)
Dropping MTU to 1380 fixed it. Wow.
Likes given · 10/12/2026
- Postgres cluster has no writable primary, all nodes in recovery
Liked by ozzie
After a region maintenance window, all three nodes of my Postgres cluster report in recovery. Nothing accepts writes. $ lark pg status -a db-prod NODE
Likes received · 10/12/2026
- Postgres cluster has no writable primary, all nodes in recovery
Liked by gilbertw
After a region maintenance window, all three nodes of my Postgres cluster report in recovery. Nothing accepts writes. $ lark pg status -a db-prod NODE
Likes received · 10/12/2026
- HTTP POST to a burrow URL does not wake a paused burrow (GET does)
Liked by deeplinker
Interesting — is the body large? Maybe the proxy gives up buffering it while the burrow wakes.
Likes received · 10/12/2026
- How to retrieve boot diagnostics for a destroyed pod?
Liked by deeplinker
An ephemeral pod failed during boot and was destroyed. lark logs has nothing because it never started the app. Are boot-stage logs kept anywhere?
Likes received · 10/12/2026
- How to retrieve boot diagnostics for a destroyed pod?
Liked by ozzie
An ephemeral pod failed during boot and was destroyed. lark logs has nothing because it never started the app. Are boot-stage logs kept anywhere?
Likes received · 10/12/2026
- How to retrieve boot diagnostics for a destroyed pod?
Liked by gilbertw
An ephemeral pod failed during boot and was destroyed. lark logs has nothing because it never started the app. Are boot-stage logs kept anywhere?
Likes received · 10/12/2026
- Blue-green deploy seems to keep adding more pods
I seem to remember that min_pods_running = 0 does not play well with blue-green: the old (blue) pods are stopped rather than destroyed, so they are not counted
Likes given · 10/12/2026
- Blue-green deploy seems to keep adding more pods
I switched one project to the bluegreen deploy strategy. I want one web pod that stops when idle and one worker that always runs. After a few deploys I had 8 we
Likes given · 10/12/2026
- Pod restarts by itself 1–3 minutes after waking from idle
Reproducible: let the app autostop, send one request, and between one and three minutes later the pod restarts. Exit code is 137. Memory graph shows nothing nea
Likes given · 10/12/2026
- New pods that mount a secret as a file are destroyed within one second
I added a file-mounted secret to lark.toml: [[files]] guest_path = "/etc/app/key.pem" secret_name = "TLS_KEY" New pods start and are destroyed within about
Likes given · 10/12/2026
- Pods can now be cordoned before a deploy
Liked by nadia_k
Does an autostopped pod stay cordoned when it wakes up?
Likes received · 10/12/2026
- The back porch: what are you running on a single shared-cpu pod?
A Minecraft server for exactly four people. It is fine until someone finds a village.
Replies · 10/12/2026
- Rails 8 solid_queue on Larkhost: one pod or two?
With solid_queue living in the same database, is it fine to run the queue inside the web process via the Puma plugin, or does it need its own pod?
Topics · 10/11/2026
- Outbound mTLS to a partner API hangs ~70 seconds then fails (iad)
70 seconds smells like an MTU problem — large client certificate, fragmented packets dropped. Try lowering the MTU on the interface and see if it changes.
Replies · 10/11/2026
- Postgres cluster has no writable primary, all nodes in recovery
After a region maintenance window, all three nodes of my Postgres cluster report in recovery. Nothing accepts writes. $ lark pg status -a db-prod NODE
Topics · 10/10/2026
- HTTP POST to a burrow URL does not wake a paused burrow (GET does)
Interesting — is the body large? Maybe the proxy gives up buffering it while the burrow wakes.
Replies · 10/10/2026