Why feature team is silent on Bulk loader performance issues?

[porsche (porscheme)]

Not sure if Bulk loader team is watching this forum, we are experiencing serious issues with bulk loader.
Appreciate if anyone chime-in to help us out.
Thanks in advance.

How to trouble shoot bulk loader hung

Version – v21.12.0 Cluster Zeros: 3 – Leader config: 32 core, 256 GB, 8 TB – Follower config: 8 Core, 64 GB, 2 TB SSD Dataset -1 – Count of Data Files: 1044 – Size of each file: 30 MB (compressed) – Result: Bulk loader hung after 36 hours Dataset -2 – Count of Data Files: 5622 – Size of each file: 30 MB (compressed) – Result: Bulk loader hung after 14 hrs Below is how we launched the bulk loader: kubectl exec -it <zero_leader_pod_name> -n K8S_namespace /bin/sh dgraph bul…

[Juri]

Zion Release is unstable atm, please use the previous release/version. Dgraph Labs is rebuilding atm the team, after that they’ll tackle that problem. Dunno how long that’ll take

[ddude]

What, are suggesting to downgrade to a year old [email protected] release?

[jdgamble555 (Jonathan Gamble)]

Yes, the year old version is stable and the only version stable enough for Dgraph Cloud.

J

[porsche (porscheme)]

I’m not sure about that.
We moved to v21.12.0 because v21.03.0 wasn’t performing…Bulk uploader hung too many times.