hexapode
commented
2 months ago We were down today and it took the status page ±45 minutes to get updated. We are working on improving our status page system so that it get updated more reliably, so we can do better in future outages.
biswaroop-p
Hello Llama parse team,
Over the past four months, I have been using llama parse to parse documents such as docx, pdf, etc... in order to convert them to markdown. Overall, my experience with llama parse has been nothing but good. However, there are times when the llama parse service is down when I cannot get my documents converted to llama pars (happens a few times a month). This in itself, is not my complaint, because outages happen, and what matters is how you respond and fix it in the future. My complaint, is that the llama-cloud status page has never accurately reflected these outages: https://llamaindex.statuspage.io/. In fact, every outage I have experienced with llama-parse, I have never seen the status page show that there was an outage. I'm guessing the primary reason for why the status page doesn't reflect outages is because it relies on the community to report outages. However, as is evident by my teams experiences, there are not enough people reporting these outages, but I don't think you need to rely on the community to report outages. Instead, a cron job that tries to have a document parsed that runs every 5 minutes or so should be able to update the status page.
Or maybe, it is simpler than that, you already track the outages, but the status page is outdated and doesn't get updated.
Either way, I would like you guys to public show the outages for llama parse as the status page never reflects this accurately.
If I am way off base here, could you please let me know why the status page for llama parse / llama cloud is not accurate?
Also, if the status page I referred to above is not for llama-parse, then the status page is misleading, and a new status page should be created with accurate results.
Thanks, Andres