Hi! I have an ongoing issue on my typesense cloud ...
# community-help
l
Hi! I have an ongoing issue on my typesense cloud instance and I feel like I’m completely in the dark about what’s happening. I did exchange one mail with support but I haven’t heard anything since then. Would it be possible to get more information about what is going on ? For information, the db startup keeps crashing and there is no way to know why.
k
Hi, I just checked on the thread. Do you index embeddings?
l
Yes, we do. We use openai-generated ones (and the server generates them for new entries automatically using an api key)
k
I think the first thing we should do is to upgrade to v29 so we can rule out some of the stability issues we identified and fixed in v28.
l
I’ll start by applying this then. Is it possible to have an idea what caused the initial crash ? Like a query of some sorts ?
k
The cause is memory spike leading to out of memory. We can also bump up the swap space temporarily to see if that helps prevent the crash and then enable query logging to isolate the query.
l
I don’t really understand what query could cause a 40+Gb spike. There’s nothing in our system that could generate such a query 🤔
k
Yes, yes, it's definitely something odd. Some run-away condition.
l
The update seems to have stabilized the system but I'm still curious about how it broke originally
k
We still see those spikes occasionally but we have now provisioned a large backup swap so that has helped. We are still investigating that. And I will get back to you.
l
Ok, I was wondering about this. Thank you for the follow-up!