since I seem to be hitting this as well (in a system with a fairly small number of accounts), intending to gradually replace nodes running swift 1.8 with nodes running swift 2.2 (adding the latter and phasing out the former), I was wondering:
do I just tolerate these errors until all the nodes are migrated (which will probably take a week or so)? or will I end up with a lost data in that case and I better do something before I proceed?
also, how does upgrading the account db schema affect remaining 1.8 nodes? I suppose they should have no problem with the upgraded schema.
since I seem to be hitting this as well (in a system with a fairly small number of accounts), intending to gradually replace nodes running swift 1.8 with nodes running swift 2.2 (adding the latter and phasing out the former), I was wondering:
do I just tolerate these errors until all the nodes are migrated (which will probably take a week or so)? or will I end up with a lost data in that case and I better do something before I proceed?
also, how does upgrading the account db schema affect remaining 1.8 nodes? I suppose they should have no problem with the upgraded schema.