Server f0a45f1e-09ee-40dd-906b-9d5ca29228c5 — slowdowns, possible shared storage contention

Posted by thashar in discord-bots.

Hello, Since yesterday I have been experiencing recurring slowdowns on my server. Server ID: f0a45f1e 09ee 40dd 906b 9d5ca29228c5 Node: PREMIUM CA EAST BHS 03 The container runs a Node.js application (/home/container). Symptoms: operations that normally take a fraction of a second — reading and writing small files, and outgoing network requests — sometimes take several seconds. It happens irregularly, a few times a day, and did not occur before. Neither the application's configuration nor its workload has changed in that time. It is not only the application itself. The panel and the console take noticeably longer to load than they did two days ago, and restarting the server also takes much longer than it used to. That happens regardless of what my application is doing, which is why I suspect the issue is on the host side rather than in my code. I understand the disk space on this node is shared, so could you please check whether there is I/O contention there — whether any container on PREMIUM CA EAST BHS 03 is generating excessive disk load, whether throttling has kicked in, and whether the underlying storage is fully healthy. I can provide exact timestamps of the occurrences if that would help. Thank you in advance for your help. Best regards, Piotr Jędryszczak

Replies

thashar

A quick follow up with additional information I gathered after my first message. My logs now show an explicit DNS error from inside the container: Error: getaddrinfo EAI AGAIN discord.com at GetAddrInfoReqWrap.onlookupall [as oncomplete] (node:dns:122:26) EAI AGAIN is a temporary failure in name resolution — at that moment the container could not resolve discord.com at all. Around the same timestamps, outgoing HTTPS requests that normally take tens of milliseconds were taking several seconds, and some failed outright. I had an AI analysis run over my logs and it suggests the root cause is DNS resolution on the node rather than anything in my application: the failures happen on the very first outgoing request, before my code does any work of its own, and they affect every part of the application equally. The same analysis notes that this is consistent with the panel and console also loading slower than usual, since that has nothing to do with my code. So in addition to the disk/IO check, could you please verify: which DNS resolvers containers on PREMIUM CA EAST BHS 03 are using, and whether they are healthy and responding within a normal time, whether there is packet loss or any network problem on that node, whether outgoing DNS traffic is being rate limited or filtered in any way. Nothing changed in my application's configuration or network usage in that period. Thank you, Piotr

Related discussions

Wispbyte Community Forum