When starting a node and it is running ActivateBestChain() in main.cpp, some RPCs hang. Currently I am seeing z_listaddresses hang but other RPC's give an error :
Why the different behavior? These are all read-only RPCs, ideally they should all be able to be run during best chain activation.
When starting a node and it is running `ActivateBestChain()` in main.cpp, some RPCs hang. Currently I am seeing `z_listaddresses` hang but other RPC's give an error :
Why the different behavior? These are all read-only RPCs, ideally they should all be able to be run during best chain activation.
```
$ ./src/hush-cli getinfo
error code: -28
error message:
Activating best chain...
$ ./src/hush-cli z_listaddresses
^C
$ ./src/hush-cli z_getoperationstatus
error code: -28
error message:
Activating best chain...
```
The RPC help is the only non-zaddr RPC I have seen hang so far, and z_getoperationstatus is the only zaddr RPC I have seen NOT hang. I have not tried them all but these clues might be enough to discover why the RPCs have different behavior.
The RPC `help` is the only non-zaddr RPC I have seen hang so far, and z_getoperationstatus is the only zaddr RPC I have seen NOT hang. I have not tried them all but these clues might be enough to discover why the RPCs have different behavior.
One difference I can see is that help uses no LOCKs while z_getoperationstatus does a LOCK on cs_main and cs_wallet . z_listaddresses uses the same LOCKs, so it's not yet clear why z_getoperationstatus and z_listaddresses behave differently.
One difference I can see is that `help` uses no LOCKs while `z_getoperationstatus` does a LOCK on `cs_main` and `cs_wallet` . `z_listaddresses` uses the same LOCKs, so it's not yet clear why `z_getoperationstatus` and `z_listaddresses` behave differently.
When starting a node and it is running
ActivateBestChain()in main.cpp, some RPCs hang. Currently I am seeingz_listaddresseshang but other RPC's give an error :Why the different behavior? These are all read-only RPCs, ideally they should all be able to be run during best chain activation.
The RPC
helpis the only non-zaddr RPC I have seen hang so far, and z_getoperationstatus is the only zaddr RPC I have seen NOT hang. I have not tried them all but these clues might be enough to discover why the RPCs have different behavior.One difference I can see is that
helpuses no LOCKs whilez_getoperationstatusdoes a LOCK oncs_mainandcs_wallet.z_listaddressesuses the same LOCKs, so it's not yet clear whyz_getoperationstatusandz_listaddressesbehave differently.