Since we have more control over the clients of wdqs in the internal cluster, that we have enough free capacity and that users are aggregated to the same user agent (mediawiki), we can (and should) be more lax than on the public wdqs cluster.
We need to experiment with the exact limits.
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| wdqs: increase throttling limits for internal cluster | operations/puppet | production | +18 -7 |
Change 465652 had a related patch set uploaded (by Gehel; owner: Gehel):
[operations/puppet@production] wdqs: increase throttling limits for internal cluster
Change 465652 merged by Gehel:
[operations/puppet@production] wdqs: increase throttling limits for internal cluster
Mentioned in SAL (#wikimedia-operations) [2018-10-11T07:57:31Z] <gehel> rolling restart blazegraph on wdqs-internal for config change - T206648