March 22, 2016 at 5:28 am
Morning all,
I'm seeing some blocking in my OLTP databases with the lead blocker's last wait type as WRITE LOG and the last command as TM REQUEST.
I've seen online that TM REQUEST can either be related to distributed transactions or Service Broker. The databases in question are Service Broker enabled but try as I might I cannot replicate this block (and I've been hammering my test system).
Has anyone else come across this?
Thanks,
Andrew
March 30, 2016 at 6:48 am
Ha ha, I'm guessing that's a no then? 🙂
If I get to the bottom of this I'll post what I find.
August 22, 2016 at 1:11 am
Ran into this problem on Friday afternoon. Very similar - WriteLog waits, then TM Request. I am using Ola's PerformanceStore and the spids associated with a slow query etc. would have a batch_statement or statement, but NONE of these spids have an associated code snippet. There are no waits for then as well. I did see, at the sametime, the following wait_types for other spids, equally affected by the problem:
PREEMPTIVE_OS_CRYPTACQUIRECONTEXT
PREEMPTIVE_OS_CRYPTIMPORTKEY
PREEMPTIVE_OS_CRYPTOPS
PREEMPTIVE_OS_DELETESECURITYCONTEXT
PREEMPTIVE_OS_WAITFORSINGLEOBJECT
I know that CPU went to 100% for a minute or so, so not sure if these are causal or symptomatic of another problem. But I would expect to see CPU-related waits along with TM Request command.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply