2022-07-06 Meeting notes

Immediate Actions from Rob C. document

All

• Deploy EL8 on Workernodes: https://stfc.atlassian.net/browse/XRD-14

• Batch farm: possible to stress test docker containers / investigate impact of packet loss (not for XrootD dev).

  • Batch experts do not think docker is a likely source (and are running latest version).

  • Do see occasional drops O(10s packets/s) (which appears much less than the number of packets succeeding a second).

• Add trace logging to dedicated WN(s) to examine failure modes.

• To check (e.g. with Rob’s test job) if there’s a time dependence to the errors.

5.4.3 releases in Centos 7 / EL8;

• Determine timescale and any blocking issues

All

https://stfc.atlassian.net/browse/XRD-14 , https://stfc.atlassian.net/browse/XRD-10

5.4.3 Potentially also solves:

• Oxford XCache (or rather 5.4.3 XCaches) (To be verified).

• Fix for how arc-ce talks to XrootD via webdav

EL8 rpms should be available;

Available stat information in XrootD

@James Walder

-xrdfs 172.28.1.1 query stats a provides potentially useful info not currently reported.

https://stfc.atlassian.net/browse/XRD-21

@Thomas Byrne

• Consider alternate mitigating options for Alice

• Spend ‘a bit' of time to understand what’s preventing the existing setup from working.

  • Come up with an ETA of any needed works.

https://stfc.atlassian.net/browse/XRD-20

@Thomas Byrne

• Demonstrated a working setup against dev-ceph

  • finished dev setup (e.g. with a proxy)

  • Replicate out to the prod gateways (xrootd gateways)