We just migrated to 15.7 SP110 version our production server :
I am wondering of some waits (extracted from asemon tool) :
The first three coumns after migrating , then Three for 9 june, and three for 15 june . I have whole day for 9 and 15 of June and half a day for 15.7.
Seems weird : amount of waitime for last log page ,
globally less waits , but much more waittime
From a time execution point of view , no specific trouble but on bad index I forgot to trap during my tests.
Is anyone who has an idea about "wait for i/o to finish after writing last log page"
and "waiting for write of the last log page to complete"
I have 9 threads and 8 disks tasks and 8 network tasks.
| ClassDesc | EventDesc | SumWaitTime(ms | Sumwaits | AvgWaitTime_ms | 09-juin | 15-juin | ||||
| waiting for a disk write to complete | wait for i/o to finish after writing last log page | 2 317 631 119 | 1 469 567 | 1 577 084,00 | 4,8 | 2,553,601 | 1.88 | 333 | 211 | 1.58 |
| waiting for input from the network | waiting for incoming network data | 1 270 324 242 | 2 748 726 | 462 150,00 | 2,343,284 | 5,333,730 | 439.33 | 10,211,897 | 428,869 | 23,811.23 |
| waiting to be scheduled | waiting on run queue after sleep | 633 441 246 | 8 599 153 | 3 663,00 | 5,97 | 32,151,248 | 0.19 | 5,796 | 33,584,659 | 0.17 |
| waiting for a disk write to complete | waiting for write of the last log page to complete | 633 437 633 | 41 845 | 15 137 713,00 | 964 | 80,95 | 11.91 | 27 | 677 | 39.88 |
| waiting for a disk read to complete | waiting for regular buffer read to complete | 417 317 522 | 1 961 527 | 212 751,00 | 48,316 | 4,433,793 | 10.90 | 15,835 | 12,317,746 | 1.29 |
| waiting for input from the network | waiting while allocating new client socket | 119 909 | 14 382 | 8 337,00 | 46,328 | 15,407 | 3,006.94 | 156,533 | 6,846 | 22,864.88 |
| waiting to be scheduled | waiting on run queue after yield | 5 164 | 2 004 909 | 2,57 | 21,116 | 4,062,346 | 5.20 | 8,945 | 1,005,109 | 8.90 |
| waiting for a disk write to complete | waiting for last i/o on MASS to complete | 1 802 | 722 957 | 2,49 | 8,917 | 1,564,100 | 5.70 | 1,715 | 389,719 | 4.40 |
| waiting for a disk write to complete | waiting for i/o on MASS initated by another task | 631 | 279 670 | 2,25 | 7,346 | 1,138,983 | 6.45 | 312 | 89,042 | 3.50 |
| waiting for a disk write to complete | waiting for buf write to complete before writing | 492 | 173 242 | 2,83 | 2,741 | 523,512 | 5.24 | 1,864 | 735,22 | 2.54 |
| waiting to output to the network | waiting for network send to complete | 375 | 133 868 | 2,80 | 4,555 | 2,861,509 | 1.59 | 137 | 160,189 | 0.86 |
| waiting to take a lock | waiting for a lock | 327 | 42 029 | 7,80 | 2,692 | 1,012,402 | 2.66 | 57 | 2,533 | 22.50 |
| waiting for memory or a buffer | waiting for MASS to finish writing before changing | 186 | 78 114 | 2,40 | 2,466 | 662,687 | 3.72 | 167 | 56,696 | 2.95 |
| waiting for input from the network | wait for data from client | 18 | 174 | 103,00 | 188 | 755 | 249.01 | 40 | 678 | 59.00 |
| waiting for memory or a buffer | waiting for buffer validation to complete | 1 | 171 | 5,80 | 19 | 973 | 19.53 | 0 | 81 | 0.00 |
| waiting for memory or a buffer | wait to write MASS while MASS is changing | 9 | 0,00 | |||||||
| waiting for memory or a buffer | waiting for buffer validation in bufnewpage | 3 | 0,00 | |||||||
| waiting for memory or a buffer | wait for access to a memory manager semaphore | 1 | 0,00 | |||||||
| waiting for memory or a buffer | wait for MASS to finish changing before changing | 63 | 0,00 | |||||||
| waiting for memory or a buffer | waiting for MASS to finish changing to start i/o | 574 | 0,00 | |||||||
| waiting for input from the network | waiting for network attn callback to complete | 1 | 0,00 | |||||||
| waiting for a disk read to complete | waiting for page reads in parallel dbcc | 13,216 | 1,838,802 | 7.19 |
The only thing which really caused a trouble was wasted space due to 8k pages migration 3 weeks ago, we previously reclaimed 30 Gb and now close to 80 Gb ...
regards
Marc