Обсуждение: BUG #14155: bloom index error with unlogged table
VGhlIGZvbGxvd2luZyBidWcgaGFzIGJlZW4gbG9nZ2VkIG9uIHRoZSB3ZWJz aXRlOgoKQnVnIHJlZmVyZW5jZTogICAgICAxNDE1NQpMb2dnZWQgYnk6ICAg ICAgICAgIFpob3UgRGlnb2FsCkVtYWlsIGFkZHJlc3M6ICAgICAgZGlnb2Fs QDEyNi5jb20KUG9zdGdyZVNRTCB2ZXJzaW9uOiA5LjZiZXRhMQpPcGVyYXRp bmcgc3lzdGVtOiAgIENlbnRPUyA2LnggeDY0CkRlc2NyaXB0aW9uOiAgICAg ICAgCgpwb3N0Z3Jlcz0jIGNyZWF0ZSB0YWJsZSB1X3RibCAoaWQgaW50KTsN CkNSRUFURSBUQUJMRQ0KcG9zdGdyZXM9IyBjcmVhdGUgaW5kZXggaWR4X3Vf dGJsIG9uIHVfdGJsIHVzaW5nIGJsb29tIChpZCk7DQpFUlJPUjogIGFjY2Vz cyBtZXRob2QgImJsb29tIiBkb2VzIG5vdCBleGlzdA0KcG9zdGdyZXM9IyBj cmVhdGUgZXh0ZW5zaW9uIGJsb29tOw0KQ1JFQVRFIEVYVEVOU0lPTg0KcG9z dGdyZXM9IyBjcmVhdGUgaW5kZXggaWR4X3VfdGJsIG9uIHVfdGJsIHVzaW5n IGJsb29tIChpZCk7DQpDUkVBVEUgSU5ERVgNCnBvc3RncmVzPSMgaW5zZXJ0 IGludG8gdV90YmwgdmFsdWVzICgxKTsNCklOU0VSVCAwIDENCnBvc3RncmVz PSMgaW5zZXJ0IGludG8gdV90YmwgdmFsdWVzICgyKTsNCklOU0VSVCAwIDEN CnBvc3RncmVzPSMgaW5zZXJ0IGludG8gdV90Ymwgc2VsZWN0IGdlbmVyYXRl X3NlcmllcygxLDEwMDAwMCk7DQpJTlNFUlQgMCAxMDAwMDANCnBvc3RncmVz PSMgZXhwbGFpbiBzZWxlY3QgKiBmcm9tIHVfdGJsIHdoZXJlIGlkPTE7DQog ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICBRVUVSWSBQTEFO ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgIA0KLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCiBHYXRoZXIgIChj b3N0PTAuMDAuLjAuMDAgcm93cz0wIHdpZHRoPTApDQogICBXb3JrZXJzIFBs YW5uZWQ6IDENCiAgIFNpbmdsZSBDb3B5OiB0cnVlDQogICAtPiAgQml0bWFw IEhlYXAgU2NhbiBvbiB1X3RibCAgKGNvc3Q9MTA0NS4zOC4uMTM5Ny40NCBy b3dzPTU2NSB3aWR0aD00KQ0KICAgICAgICAgUmVjaGVjayBDb25kOiAoaWQg PSAxKQ0KICAgICAgICAgLT4gIEJpdG1hcCBJbmRleCBTY2FuIG9uIGlkeF91 X3RibCAgKGNvc3Q9MC4wMC4uMTA0NS4yNCByb3dzPTU2NQp3aWR0aD0wKQ0K ICAgICAgICAgICAgICAgSW5kZXggQ29uZDogKGlkID0gMSkNCig3IHJvd3Mp DQoNCnBvc3RncmVzPSMgc2VsZWN0ICogZnJvbSB1X3RibCB3aGVyZSBpZD0x Ow0KIGlkIA0KLS0tLQ0KICAxDQogIDENCigyIHJvd3MpDQoNCnBvc3RncmVz PSMgZHJvcCB0YWJsZSB1X3RibCA7DQpEUk9QIFRBQkxFDQpwb3N0Z3Jlcz0j IGNyZWF0ZSB1bmxvZ2dlZCB0YWJsZSB1X3RibCAoaWQgaW50KTsNCkNSRUFU RSBUQUJMRQ0KcG9zdGdyZXM9IyBjcmVhdGUgaW5kZXggaWR4X3VfdGJsIG9u IHVfdGJsIHVzaW5nIGJsb29tIChpZCk7DQpFUlJPUjogIGluZGV4ICJpZHhf dV90YmwiIGFscmVhZHkgY29udGFpbnMgZGF0YQ0KcG9zdGdyZXM9IyBcc2V0 IFZFUkJPU0lUWSB2ZXJib3NlDQpwb3N0Z3Jlcz0jIGNyZWF0ZSBpbmRleCBp ZHhfdV90Ymwgb24gdV90YmwgdXNpbmcgYmxvb20gKGlkKTsNCkVSUk9SOiAg WFgwMDA6IGluZGV4ICJpZHhfdV90YmwiIGFscmVhZHkgY29udGFpbnMgZGF0 YQ0KTE9DQVRJT046ICBibGJ1aWxkZW1wdHksIGJsaW5zZXJ0LmM6MTY0DQpw b3N0Z3Jlcz0jIGRyb3AgdGFibGUgdV90YmwgOw0KRFJPUCBUQUJMRQ0KcG9z dGdyZXM9IyBjcmVhdGUgdGVtcCB0YWJsZSB1X3RibCAoaWQgaW50KTsNCkNS RUFURSBUQUJMRQ0KcG9zdGdyZXM9IyBjcmVhdGUgaW5kZXggaWR4X3VfdGJs IG9uIHVfdGJsIHVzaW5nIGJsb29tIChpZCk7DQpDUkVBVEUgSU5ERVgNCnBv c3RncmVzPSMgaW5zZXJ0IGludG8gdV90Ymwgc2VsZWN0IGdlbmVyYXRlX3Nl cmllcygxLDEwMDAwMCk7DQpJTlNFUlQgMCAxMDAwMDANCnBvc3RncmVzPSMg ZXhwbGFpbiBzZWxlY3QgKiBmcm9tIHVfdGJsIHdoZXJlIGlkPTE7DQogICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICBRVUVSWSBQTEFOICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgCg0KLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0NCiBCaXRtYXAgSGVhcCBTY2FuIG9uIHVfdGJs ICAoY29zdD0xMDQ1LjM4Li4xMzk3LjQ0IHJvd3M9NTY1IHdpZHRoPTQpDQog ICBSZWNoZWNrIENvbmQ6IChpZCA9IDEpDQogICAtPiAgQml0bWFwIEluZGV4 IFNjYW4gb24gaWR4X3VfdGJsICAoY29zdD0wLjAwLi4xMDQ1LjI0IHJvd3M9 NTY1CndpZHRoPTApDQogICAgICAgICBJbmRleCBDb25kOiAoaWQgPSAxKQ0K KDQgcm93cykNCg0KcG9zdGdyZXM9IyBzZWxlY3QgKiBmcm9tIHVfdGJsIHdo ZXJlIGlkPTE7DQogaWQgDQotLS0tDQogIDENCigxIHJvdykKCg==
digoal@126.com writes:
> postgres=# create unlogged table u_tbl (id int);
> CREATE TABLE
> postgres=# create index idx_u_tbl on u_tbl using bloom (id);
> ERROR: index "idx_u_tbl" already contains data
Yeah, it looks like nobody ever tested bloom's unlogged-index support;
it doesn't work or even come very close to working. Will fix, thanks
for the report!
regards, tom lane
On Tue, May 24, 2016 at 5:08 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote: > digoal@126.com writes: >> postgres=# create unlogged table u_tbl (id int); >> CREATE TABLE >> postgres=# create index idx_u_tbl on u_tbl using bloom (id); >> ERROR: index "idx_u_tbl" already contains data > > Yeah, it looks like nobody ever tested bloom's unlogged-index support; > it doesn't work or even come very close to working. Will fix, thanks > for the report! To be honest, I began hacking around that to address the issue, though I don't have yet something to post... So I guess that there is nothing I need to do? Or do you need a patch? -- Michael
On Tue, May 24, 2016 at 6:04 PM, Michael Paquier <michael.paquier@gmail.com> wrote: > On Tue, May 24, 2016 at 5:08 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote: >> digoal@126.com writes: >>> postgres=# create unlogged table u_tbl (id int); >>> CREATE TABLE >>> postgres=# create index idx_u_tbl on u_tbl using bloom (id); >>> ERROR: index "idx_u_tbl" already contains data >> >> Yeah, it looks like nobody ever tested bloom's unlogged-index support; >> it doesn't work or even come very close to working. Will fix, thanks >> for the report! > > To be honest, I began hacking around that to address the issue, though > I don't have yet something to post... So I guess that there is nothing > I need to do? Or do you need a patch? Or too late... -- Michael
Michael Paquier <michael.paquier@gmail.com> writes:
>> Yeah, it looks like nobody ever tested bloom's unlogged-index support;
>> it doesn't work or even come very close to working. Will fix, thanks
>> for the report!
> To be honest, I began hacking around that to address the issue, though
> I don't have yet something to post... So I guess that there is nothing
> I need to do? Or do you need a patch?
Nah, I just pushed something ... but if you'd like to go over the
documentation thread David was complaining about and produce a cleaned-up
docs patch, that would be pretty helpful.
regards, tom lane
On Tue, May 24, 2016 at 6:11 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote: > Michael Paquier <michael.paquier@gmail.com> writes: >>> Yeah, it looks like nobody ever tested bloom's unlogged-index support; >>> it doesn't work or even come very close to working. Will fix, thanks >>> for the report! > >> To be honest, I began hacking around that to address the issue, though >> I don't have yet something to post... So I guess that there is nothing >> I need to do? Or do you need a patch? > > Nah, I just pushed something ... but if you'd like to go over the > documentation thread David was complaining about and produce a cleaned-up > docs patch, that would be pretty helpful. OK. I'll get a look at this thread and see how the documentation can be improved by replying there, though there is already a patch I see... -- Michael