\uD83D\uDDD3 Date
\uD83D\uDC65 Participants
Lancs: Gerard, Steven
Glas: Sam
\uD83E\uDD45 Goals
List of Epics
New tickets
Consider new functionality / items
Detailed discussion of important topics
Site report activity
\uD83D\uDDE3 Discussion topics
...
Item | Presenter | Notes | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
lockless Reads status update | Last Friday latest version rolled out to WNs. No more locks; remaining sets of metadata ops taking a greater proportion of original metadata time. Fewer unlocks overall, but seem to be slower? To follow up on the unlocks; source of the slow unlocks. How to reduce the unlock operations. For unlocks: still looking into the completion objects. | |||||||||||||||||
5.4.3 releases in Centos 7 / EL8;
| All |
5.4.0 enabled by default pgRead and pgWrite on clients. Significant speed reduction. (due to 64kib r/w size). Disabling pgRW requires modifying the XRD_PGProtocolversion (a hard-coded #define), so that the client falls back to a standard Read/Write. Test against the bufferedIO version also needed. Tom indicates | ||||||||||||||||
Deletions Status | Ian Johnson , all | To look at 5,10,100GB deletes Other thoughts on unified xrootd instances and possible changes to ceph.conf ops | ||||||||||||||||
Deletions Strategy | all | Postponed | ||||||||||||||||
Site reports
Lancaster
Matt (in absentia) - Daniela spotted that xrdcp with our xroot server as a destination is broken. I’ve tried a few different incantations but no joy - currently just commented out all so at least the error message is clear.
#original
#ofs.tpc cksum adler32 fcreds ?gsi =X509_USER_PROXY autorm xfr 40 pgm /etc/xrootd/xrdcp-tpc.sh
#no checksum
#ofs.tpc fcreds ?gsi =X509_USER_PROXY ttl 60 70 xfr 100 autorm pgm /etc/xrootd/xrdcp-tpc.sh
#xrdcp
#ofs.tpc fcreds ?gsi =X509_USER_PROXY ttl 60 70 xfr 100 autorm pgm /usr/bin/xrdcp