Pedro Lopes comments

Results 13 comments of


                                            Pedro Lopes

AdaptiveIndexDefrag - bug with ForceRescan Option - lines 1732 - 1750

HI, Apologies for taking this long to take a look. About the compression_type, I see the problems. Actually removed the upgrade condition - if anyone is upgrading from such an...

Broken subsection title

Thanks this is some weird artifact. Fixing it now.

no rebuild

Apologies for the delay - Github was not sending updates. It would be useful to get a repro. Can you run with @Exec_Print = 0, @printCmds=0 and see if it...

usp_whatsup @qrystats=1 throwing error

Should be fixed in the latest upload. Thanks for reporting.

usp_whatsup @qrystats=1 throwing error

I'm sorry but I cannot repro... Character 0x0000 is NCHAR(0) and it's escaped in the REPLACE function. It would help if you can dump your DMV into a table and...

usp_whatsup @qrystats=1 throwing error

Thanks for that, but it's not the semicolon. Coming from input buffer or exec_sql_text is not the cause, rather the XML formatting encountering an illegal char. And if running in...

[AdaptiveIndexDefrag] Index_level filter conflict with @scanMode = DETAILED

So we're targeting leaf level fragmentation only which is why its filtering on 0, and then affects subsequent logic. For 0 fragmentation of the leaf-level, it's curious you had such...

[AdaptiveIndexDefrag] null @DataCompression parameter leads to null @rebuildCommand

HI, should be fixed with #94 just checked-in for v1.6.6.6. Can you please verify?

Update usp_AdaptiveIndexDefrag.sql

There shouldn't be a NULL stats id. Can you please turn on debug mode and let me know where this error is triggered? Maybe even make available the db clone...

Columnstore fragmentation can be > 100, but @rebuildThreshold_cs is capped at 100

There's no logical point is using a percent above 100. Capped the max columnstore frag calculation to 100 in v1.6.6.8. If I'm not following your intent, please explain.