Error: 7884, Severity: 20, State: 1. (Params:). The error is printed in terse mode because there was error during formatting.

Tilicho 6

Error seen:
Error: 7884, Severity: 20, State: 1. (Params:). The error is printed in terse mode because there was error during formatting.

This is for 2014 SQL server with alwaysOn group that hosts databases for Sharepoint. All are virtual machines on esxi. Over a week back this noticed popped up (the server has been in production for about two years and first time witnessing the error). Ran consistency check on the databases and no errors.

The error popped up again after a week yesterday. Ran checksum and no errors. And again this afternoon.

I have gone through numerous links in the web and none seem to be helpful. There are no errors in sql log prior to or after that one particular error shows. No error in event viewer either during that time.

Looking at sessions, one time the session id seem to point the web front end of sharepoint as source with secure store database as backend. Second time the session id showed the application server as source that was using the search database.

https://social.msdn.microsoft.com/Forums/en-US/5d9a6149-043b-4fed-9909-078b7b870239/the-error-is-printed-in-terse-mode-because-there-was-error-during-formatting?forum=sqldatabaseengine

wasn't helpful in that it talks about sql profiler (since the error is sporadic running profiler forever doesn't help) and also the microsoft connect sites some of these articles references are no longer in use. I have "googled" and not found anything substantial.

Any tips or recommendations would be appreciated.

Erland Sommarskog 107.2K Reputation points

2021-09-14T21:14:03.06+00:00

Where do you see this error? Have you looked in the SQL Server errorlog?

What is the output of "SELECT @@version"?
Tilicho 6 Reputation points

2021-09-14T21:28:41.487+00:00

The error is from the SQL Server Logs. Nothing in the SQL Server Agent error logs. And nothing in windows applications, security, or system logs around the time that the error in sql server logs showed.

Version is sql 2014 with SP3.
Tilicho 6 Reputation points

2021-09-21T19:25:22.8+00:00

@Erland Sommarskog @Tom Phillips @AmeliaGu-MSFT

Lowered the max memory in primary replica (alwayon group has two replicas. The secondary had a limit set and primary didn't. So had primary match up with secondary). Some time after the change got the same error which I ignored since the server hadn't been rebooted.

Servers were rebooted over weekend. Encountered two episodes of the errors again today. The extended events points to the sharepoint portal content database with client being the sharepoint webserver1. The other time it was for the same database but from sharepoint webserver2. Both cases ties the username to the application pool service account. It has a lengthy sql text and look similar in both incidents.

There are no errors right before this happens. Right after though is followed by message of transaction was deadlocked on lock resources with another process and has been chosen as the deadlock victim. The victim is the process id that generates the 7884 error.

3 answers

Erland Sommarskog 107.2K Reputation points

2021-09-14T21:32:37.2+00:00

Rather than using Profiler, you can use an Extended Events session. This is useful so far, that you can capture only the even error_reported, but with this event also get the statement which produced this error. This is difficult to achieve with Trace/Profiler.

I have an article on my web site on how to set up such an event session and how to query it: https://www.sommarskog.se/Short%20Stories/trace-exceptions.html.
Please sign in to rate this answer.
Tilicho 6 Reputation points

2021-09-14T21:51:29.88+00:00

Thank you! The blog was helpful in setting up the extended event. I have "error_reported" selected with "client_app_name, client_hostname, database_name, server_principal_name, session_id, session_nt_username, sql_text, tsql_frame, tsql_stack, username" selected for global fields. The data storage is set to save to file with max size 1 GB with rollover.

Hopefully, it helps to catch the culprit. I will update after the error pops up again.

AmeliaGu-MSFT 13,971 Reputation points Microsoft Vendor

2021-09-17T09:32:43.113+00:00

Hi Tilicho-7232,

How are things going?
Have you captured any exceptions?
Please feel free to let us know if you have any other question. If you find any post in the thread is helpful, you could kindly accept it as answer.

Best Regards,
Amelia

Tilicho 6 Reputation points

2021-09-21T19:27:44.703+00:00

@Erland Sommarskog @Tom Phillips @AmeliaGu-MSFT

Lowered the max memory in primary replica (alwayon group has two replicas. The secondary had a limit set and primary didn't. So had primary match up with secondary). Some time after the change got the same error which I ignored since the server hadn't been rebooted.

Servers were rebooted over weekend. Encountered two episodes of the errors again today. The extended events points to the sharepoint portal content database with client being the sharepoint webserver1. The other time it was for the same database but from sharepoint webserver2. Both cases ties the username to the application pool service account. It has a lengthy sql text and look similar in both incidents.

There are no errors right before this happens. Right after though is followed by message of trasaction was deadlocked on lock resources with another process and has been chosen as the deadlock victim. The victim is the process id that generates the 7884 error.

Erland Sommarskog 107.2K Reputation points

2021-09-21T21:07:37.707+00:00

Hm, could there be an issue with writing the deadlock trace? Are you running with trace flag 1222 active?

I should hasten to add that this is wild speculation on my part.

Tilicho 6 Reputation points

2021-09-22T13:29:32.713+00:00

@Erland Sommarskog @AmeliaGu-MSFT @Tom Phillips

Thank you for response.

The extended sessions do show other deadlocks events (error 1205) but does not have the 7884 beforehand. The deadlocks errors seen in extended session don't show up in server logs.

The last few times when this happened I have run dbcc checkdb with 0 errors. Not sure what is up. Seeing this pop up infrequently and first time since the servers went in production about a year and a half ago.

Erland Sommarskog 107.2K Reputation points

2021-09-22T21:18:04.39+00:00

The extended sessions do show other deadlocks events (error 1205) but does not have the 7884 beforehand. The deadlocks errors seen in extended session don't show up in server logs.

So there are no deadlock errors in the SQL Server errorlog? That is perfectly normal, because they are only printed if TF 1222 is active.

Or are you saying that you see output for the deadlocks that are not preceded by 7884?

Tilicho 6 Reputation points

2021-09-23T13:26:15.68+00:00

@Erland Sommarskog @AmeliaGu-MSFT

No, there are no deadlock errors in the sql server log. In the sql server log only error reported is for 7884. This particular error is not frequent but has been occurring sporadically for about three weeks now.

The deadlock errors can be seen in extended events sessions. When 7884 occurs I see a deadlock for the spid that is associated with 7884 which then gets killed. For other deadlocks that I see in the session file I do not see any 7884 association.
Sign in to comment
Tom Phillips 17,721 Reputation points

2021-09-15T11:49:35.863+00:00

This kind of error is almost always related to a memory allocation error. I would try reducing your max server memory setting a little and see if you get a real error message.
Please sign in to rate this answer.

0 comments No comments
Sign in to comment
Erland Sommarskog 107.2K Reputation points

2021-09-23T21:42:39.437+00:00

I don't think we come any further with this. If you want to know more, I think you need to open a support case. Which means that you must have purchased extended support, since you are on SQL 2014. To me, this seems like something internal in SQL Server.

Then again, beside the mystery, have you any real problems with things not working properly?
Please sign in to rate this answer.
Tilicho 6 Reputation points

2021-09-23T22:04:11.217+00:00

No support was purchased. I think the norm is to open a ticket with MS and pay on ticket basis. I doubt that will be approved.

Besides the mystery, no issues that I am aware of. I am not a sharepoint person and tagged others here from Sharepoint community hoping they would be able to help. I have checked the databases for consistency and returned 0 errors.

Received two more errors this morning. It is starting to get more frequent now than it was before in that I received the second occurrence of the error after a week of the first one. Now it is like every few days or two days in a row (yesterday and today).

Erland Sommarskog 107.2K Reputation points

2021-09-23T22:06:47.667+00:00

I will have to admit that it gives an uneasy feeling. Nothing is real problem yet, but nay be the house is about to crumble totally?
Sign in to comment

Share via

Error: 7884, Severity: 20, State: 1. (Params:). The error is printed in terse mode because there was error during formatting.

3 answers