Hosting troubles in Azure-UK, no log entries after server started #152
-
|
I've read through the two other Azure related discussions #120 and #137, and we're getting different behaviour from either of those two. We're able to download and begin execution of the server without issue. Unfortunately the server seems to go unresponsive after "starting". Here's the container output from Azure, at 16:03 UTC The cron job is running regularly to check for updates without issue, and we're seeing messages from it just fine. For reference here's our container json. There's missing environment variables for password and port, but the former was created with a {
"properties": {
"provisioningState": "Succeeded",
"containers": [
{
"name": "ct-valheim-podlings",
"properties": {
"image": "lloesche/valheim-server",
"ports": [
{
"protocol": "UDP",
"port": 2456
},
{
"protocol": "UDP",
"port": 2457
},
{
"protocol": "UDP",
"port": 2458
}
],
"environmentVariables": [
{
"name": "SERVER_NAME",
"value": "podlings"
},
{
"name": "WORLD_NAME",
"value": "Dedicated_pods"
}
],
"instanceView": {
"restartCount": 1,
"currentState": {
"state": "Running",
"startTime": "2021-03-03T15:52:46Z",
"detailStatus": ""
},
"previousState": {
"state": "Terminated",
"startTime": "2021-03-03T15:33:32Z",
"exitCode": 137,
"finishTime": "2021-03-03T15:52:40Z",
"detailStatus": "OOMKilled"
},
"events": [
{
"count": 2,
"firstTimestamp": "2021-03-03T15:32:41Z",
"lastTimestamp": "2021-03-03T15:52:43Z",
"name": "Pulling",
"message": "pulling image \"lloesche/valheim-server\"",
"type": "Normal"
},
{
"count": 2,
"firstTimestamp": "2021-03-03T15:33:23Z",
"lastTimestamp": "2021-03-03T15:52:46Z",
"name": "Pulled",
"message": "Successfully pulled image \"lloesche/valheim-server\"",
"type": "Normal"
},
{
"count": 2,
"firstTimestamp": "2021-03-03T15:33:32Z",
"lastTimestamp": "2021-03-03T15:52:46Z",
"name": "Created",
"message": "Created container",
"type": "Normal"
},
{
"count": 2,
"firstTimestamp": "2021-03-03T15:33:32Z",
"lastTimestamp": "2021-03-03T15:52:46Z",
"name": "Started",
"message": "Started container",
"type": "Normal"
},
{
"count": 1,
"firstTimestamp": "2021-03-03T15:52:40Z",
"lastTimestamp": "2021-03-03T15:52:40Z",
"name": "Killing",
"message": "Killing container with id docker://ct-valheim-podlings:Need to kill Pod",
"type": "Normal"
}
]
},
"resources": {
"requests": {
"memoryInGB": 1.5,
"cpu": 1
}
},
"volumeMounts": [
{
"name": "azurefile",
"mountPath": "/config"
}
]
}
}
],
"restartPolicy": "Always",
"ipAddress": {
"ports": [
{
"protocol": "UDP",
"port": 2456
},
{
"protocol": "UDP",
"port": 2457
},
{
"protocol": "UDP",
"port": 2458
}
],
"ip": "20.77.146.186",
"type": "Public",
"dnsNameLabel": "valheim-podlings",
"fqdn": "valheim-podlings.uksouth.azurecontainer.io"
},
"osType": "Linux",
"volumes": [
{
"name": "azurefile",
"azureFile": {
"shareName": "fs-valheim-pod",
"storageAccountName": "savalheimpodlings"
}
}
],
"instanceView": {
"events": [
{
"count": 1,
"firstTimestamp": "2021-03-03T15:32:40Z",
"lastTimestamp": "2021-03-03T15:32:40Z",
"name": "SuccessfulMountAzureFileVolume",
"message": "Successfully mounted Azure File Volume.",
"type": "Normal"
}
],
"state": "Running"
}
},
"id": "/subscriptions/283b862a-c701-4cff-8143-b7426cc94c8e/resourceGroups/RG-Valheim-podlings/providers/Microsoft.ContainerInstance/containerGroups/ct-valheim-podlings",
"name": "ct-valheim-podlings",
"type": "Microsoft.ContainerInstance/containerGroups",
"location": "uksouth",
"tags": {}
}any insight you can offer would be greatly appreciated! We're quite stumped. |
Beta Was this translation helpful? Give feedback.
Replies: 4 comments 4 replies
-
As I am building my own docker valheim server image right now, i can only relate to this behavior. I was getting this "disconnected" message when my server and client version where not matching. |
Beta Was this translation helpful? Give feedback.
-
|
Hmm, will double check but I'm fairly sure both the server and the clients were latest version vanilla |
Beta Was this translation helpful? Give feedback.
-
|
Regarding the disconnected, verify that the ACI IP address has not changed since you initially set it. I updated a few things and restarted it a few times which rolled the IP without me realizing causing us to get "disconnected" when trying to join (since no server was actually being hosted at that IP anymore). You can avoid this by using the ACI FQDN instead of the IP to join the game as it supports DNS. Regarding the logs stopping abruptly, I am also seeing similar issues. |
Beta Was this translation helpful? Give feedback.
-
|
To update on this - it seems that the Azure hosting just takes a /really/ long time to start up (around 15 minutes). It's unclear if the CRON job to check for updates is causing issues here, but my next steps of troubleshooting will be to push the frequency of that process back to around once an hour. Will update if that provides and useful progress. |
Beta Was this translation helpful? Give feedback.
-
|
@Ctri-The-Third the update check won't interfere with the ongoing startup unless there is an actual update to be downloaded. In which case it'll perform the update and restart the (not yet started) server. So worst case the startup begins anew. |
Beta Was this translation helpful? Give feedback.
-
That's not even close to enough resources. Try a minimum of 2 cores and 4 GB. But realistically 4 cores and 16 GB if you don't want to swap. Our server consumes 3GB RSS and 11GB VSZ. |
Beta Was this translation helpful? Give feedback.
-
|
Actually, does Azure even support swapping? If not then I'd try with 12 GB and keep an eye on memory consumption. CPU wise the server creates around 50 threads but only two or three have any real load on them. Most load is handled by a single thread. Still I'd go with at least two cores. |
Beta Was this translation helpful? Give feedback.
-
|
That explains it! ^^; |
Beta Was this translation helpful? Give feedback.
That's not even close to enough resources. Try a minimum of 2 cores and 4 GB. But realistically 4 cores and 16 GB if you don't want to swap. Our server consumes 3GB RSS and 11GB VSZ.