network profile
| bio | website | spartansql.com |
|---|---|---|
| location | Minnesota | |
| age | ||
| visits | member for | 2 years, 4 months |
| seen | 14 hours ago | |
| stats | profile views | 14 |
A SQL Server DBA living and working in the Twin Cities. Professional interests include replication and T-SQL programming.
|
May 8 |
comment |
How to find when last reindexing happened in sql server? @MichaelJSwart: The question was more rhetorical than anything. The answers you gave are good ones! But answering a question without knowing the why is often providing disservice to the asker. |
|
May 2 |
comment |
Restoring differential backup from SQL Server 2005 to SQL Server 2012 Did anyone take a full backup between the full backup that you restored and the differential backup that you tried (and failed) to restore? Check msdb.dbo.backuphistory. |
|
Apr 30 |
answered | Create SQL Script for Statistics |
|
Apr 15 |
comment |
Service Broker - Communication not happening between servers despite ssbdiagnose saying all is well Could be, though that would be strange to me that it would work in the homogeneous situation and not the heterogeneous. If it were me, I'd give it a shot. It's really unlikely to break anything and if it doesn't fix the problem, you can do alter master key drop encryption by service master key |
|
Apr 12 |
comment |
Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber let us continue this discussion in chat |
|
Apr 12 |
answered | Service Broker - Communication not happening between servers despite ssbdiagnose saying all is well |
|
Apr 12 |
comment |
Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber I logged a connect item a while back for this very issue (same issue which causes me to know that you need to be subscribed to the first article to get tokens to commit at the subscriber). The response was that it's an unsupported configuration to not subscribe to all articles in a pub. So my advice: either get rid of the unsubscribed to articles, subscribe to them, or move them to their own publication. |
|
Apr 12 |
comment |
Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber Don't get me wrong; that wasn't meant as anything towards you personally. People just get intimidated because "replication is hard". So you're seeing the commands being applied at the subscriber. Tracer tokens are just marks made in the transaction log at the publisher that the log reader and distribution agents process. Dumb question, but are you subscribing to all articles in the publication? IIRC, tokens don't get processed if you don't subscribe to the first article in the publication. |
|
Apr 12 |
comment |
Windows native backups makes SQL Server think that databases backup has been done What do the various backup* tables (i.e. backupset, backupmediafamily, backupmediaset) say about it? Incidentally, the restore that you were "forced" to perform doesn't sound odd to me given how you've set up your backups. What restore did you try that didn't work as intended? |
|
Apr 11 |
comment |
Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber One other thing to keep in mind; there is a clock ticking. Specifically, the minimum retention property on the distribution database is when repl'n is allowed to start trimming. So you may want to increase that while you troubleshoot. |
|
Apr 11 |
comment |
Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber People think replication is magic. It's not; it's just another application making inserts, updates, and deletes (through stored procedures) in your database. So treat it as such. Is there blocking? A bad query plan (e.g. parameter sniffing) for one or more of the procs? Fire up a trace and see what's happening. That all assumes that there aren't any errors as far as the repl'n monitor is concerned and that the distribution agent is running... |
|
Apr 11 |
answered | Replication Monitor reports little to no latency, but my tracer token is taking hours to go from Distributor to Subscriber |
|
Apr 8 |
comment |
Why is CDC capture failing due to a missing dbo principal Under what account does your SQL Agent run? My guess is that it's the NT AUTHORITY\NETWORK SERVICE you mentioned. I'd suggest changing that to an explicit account. Failing that, add that account to the db_owner role in the database. |
|
Mar 4 |
asked | Are “create user” and “grant connect” functionally equivalent? |
|
Feb 22 |
comment |
Database mirroring - link failure between principal and witness Not to agree or disagree with you, but the witness has an option on how to make the situation healthy. That is if it voted that the mirror was able to take the load, the system would report as "healthy" again. Do you have corroborating evidence that it works the way you say it works? |
|
Feb 6 |
comment |
How to find when last reindexing happened in sql server? My question: why do you care? An index on a table with no writes that was "re-indexed" three years ago will still be unfragmented. Conversely, an index that encounters a lot of page splits in a short period of time could be heavily fragmented in five minutes. |
|
Jan 31 |
comment |
SQL Server replication subscriptions marked as inactive I'd try setting your min_distretention higher than 0. You didn't mention what your max was, but I suspect that the cleanup agent is cleaning up commands before they have a chance to get replicated to your disconnected subscriber. I typically set it to a day (i.e. 24, since min_distretention is in hours). |
|
Jan 30 |
comment |
SQL Server replication subscriptions marked as inactive I don't think 0 means infinite here. Check out blogs.technet.com/b/claudia_silva/archive/2009/05/04/…. Either way, you're not storing them for an infinite amount of time. How big would your distribution database be if you were? |
|
Jan 29 |
comment |
SQL Server replication subscriptions marked as inactive I've never had to re-init a subscription where there were connection issues that were resolved within the distribution database's minimum retention period. What is that setting for you? |
|
Jan 10 |
awarded | Yearling |
