presently my vp2 and gw1000 data are merged in one database, and i am splitting them. i took the weewx.sdb and placed it in new weewx 4.7.0 environment. i ran weewx a couple of times configured to create empty vp2.sdb and gw1000.sdb (with vp2.conf and gw1000.conf respectively pointing to their own schemas and dbs). i then copied the relevant observations (own program) from weewx.sdb columns into vp2.sdb and gw1000.sdb then ran ‘wee_database —conf=__.conf —calc-missing’ to fill in the respective derived values. the vp2 (with 16 years of data) ran in about 10 mins, as expected. the gw1000 (with 2 years of data) has been running for 12 hours so far, at about 7000 recs/hour, and at present rate will take about 2 days in total, which ain’t gonna fly when i try actual migration of production environment. note: when processing the vp2 data, cpu is typically 98%, but with gw1000 data wee_database dawdles on 0.3% and usually lower.
before i dive into detailed analysis on why —calc-missing is orders of magnitude slower on the gw1000 data than the vp2 data, can someone provide me some pointers, guidance where to look? (if this doesn’t work, i’ll try again by moving columns around copies of weewx.sdb rather than programmatically splitting the data, and see if i get —calc-missing issue again…) cheers ⊣ Graham Eddy ⊢ -- You received this message because you are subscribed to the Google Groups "weewx-user" group. To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/CD5AE3A9-7A5B-4A1E-8D63-08B537D4C259%40gmail.com.
