Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Wed, 2006-06-28 at 10:36 +0200, Lars Ellenberg wrote: > / 2006-06-28 12:42:00 +0530 > \ Milind Dumbare: > > Hi, > > I just found out that we can do this by using coomand line options of > > drbdsetup. > > #drbdsetup /dev/drbd0 syncer --rate=5M --skip-sync > > this is a sticky option introduced only to more easily reproduce certain > corner cases during development and debugging. you won't get _any_ sync > until you reset this on both nodes while they are not connected, then > reconnect (and then you will get the outstanding sync). > > > Thanks > > (Wrote answer for archive purpose) > > stay away from it. > it is not for "testing". > it will not help you during testing. > it will cause problems and strange behaviour. > it is for developers only, so we can more > conveniently debug certain corner cases. > stay away from it. > Thanks a lot for your help > if you want to skip the _initial_ sync, you can use the "write_gc.pl" > (testing/ subdirectory of the tarball) to initialize the generation > counts to something identical and consistent before you use the drbd > peers for the first time. ok, so write_gc.pl initializes the generation counts of two partition two identical values write? which avoids initial sync, right? So will following command #write_gc.pl resource-name [--del-al] counter-values initialize both (primary and secondary) underlying device's generation count to same value? Thanks again for the warning. > there are threads about avoiding the initial > sync in the archives. > > stay away from the option you propose to use. > stay away from it. > > thanks. > -- ...Milind KGDB-PRO http://www.linsyssoft.com/Kgdbpro.html