On our Gatsby+WordPress site uwalumni (2ff51eee-cdcc-4ab6-92d2-b5d2eac9190e), I have been struggling to get successful builds going on the branch “netlify”, where I’ve been making a few Netlify-specific config changes. Last night two builds failed during the deployment phase, after the build phase had successfully completed. They both showed something like this in the logs during the deployment phase:
10:42:18 PM: Deploy site
10:42:18 PM: ────────────────────────────────────────────────────────────────
10:42:18 PM:
10:42:19 PM: Starting to deploy site from “public”
10:42:50 PM: Calculating files to upload
10:43:11 PM: 7832 new files to upload
10:43:11 PM: 0 new functions to upload
10:43:11 PM: Starting to upload
10:43:16 PM: 10% uploaded
10:43:22 PM: 20% uploaded
10:43:25 PM: 30% uploaded
10:43:30 PM: 40% uploaded
10:43:36 PM: 50% uploaded
10:43:41 PM: 60% uploaded
10:43:47 PM: 70% uploaded
10:43:51 PM: 80% uploaded
10:43:55 PM: 90% uploaded
10:43:59 PM: 100% uploaded
10:43:59 PM: Section completed: deploying
10:44:05 PM: Finished waiting for live deploy in 5.858s
10:44:05 PM:
10:44:05 PM: Configuration error
10:44:05 PM: ────────────────────────────────────────────────────────────────
10:44:05 PM:
10:44:05 PM: Error message
10:44:05 PM: Deploy did not succeed: timed out while waiting to enter states [ready]
10:44:05 PM:
And in the list of deploys, the summary message for these failed deploys says
"The document exceeds maximum allowed BSON object size after serialization”
There are some additional log entries with page-specific headers:
headers:
10:44:05 PM: - for: /page-data/groups/bayarea/about/support/store/page-data.json
10:44:05 PM: values:
10:44:05 PM: Referrer-Policy: strict-origin-when-cross-origin
10:44:05 PM: cache-control: public, max-age=0, must-revalidate
10:44:05 PM: referrer-policy: same-origin
10:44:05 PM: x-content-type-options: nosniff
10:44:05 PM: x-frame-options: DENY
10:44:05 PM: x-xss-protection: 1; mode=block
…etc
I do recall seeing a message having something to do with headers being too large a few weeks ago, and I’m wondering if that has something to do with this. But I can find no info about this message, and these deploys don’t correspond to any significant code changes.
These failures both happened at about 78 minutes from the start time.
The next failed deploy after these two was a timeout at 93 minutes.