By using AWS re:Post, you agree to the Terms of Use

Assert Error in redshift

0

In the beginning of the month, we changed our Redshift-instance from virginia to Sao Paulo (sa-east). Since then, we are receiving several times such error:

ERROR: Assert
Detail:

error: Assert
code: 1000
context: p - id=492 not in rowsets list
query: 321640
location: tbl_trans.cpp:412
process: query0_75 pid=12577

We use Looker as Business Intelligence software, which is connected to redshift, and this error occurs when we try to schedule a report by e-mail.
I already checked in forums, and I didn't find a question with the same context ("not in rowsets list").

How can we circumvent this?

asked 5 years ago1096 views
19 Answers
0
Accepted Answer

Thanks, Jeff. The release containing the fix for this assert error is now available in all regions (as of yesterday afternoon). Appreciate if you can confirm that after upgrading to a release version >= 1.0.1636, the issue is resolved.

answered 5 years ago
0

Thank you for bringing this to our attention.

Can you confirm if retrying the query allows it to succeed?

answered 5 years ago
0

Hello,
Same error here using BIME a Business Intelligence report software.
The strange thing is that this error started occurred recently ~1 week especially when we add filters to the query.
A simple re-run of the query seems to fix it but it's quite a problem since the error appears randomly.

answered 5 years ago
0

Hello @joeharris76-AWS ,

It is exactly what RonanTreillet answered:

"A simple re-run of the query seems to fix it but it's quite a problem since the error appears randomly."

answered 5 years ago
0

Sorry you've experienced this problem. We have identified the issue and are working on a fix.

Please look for the fix to appear in our biweekly maintenance announcements at the top of the forum.

answered 5 years ago
0

Hello,
Are you still working on the fix? The problem listed above remains, and I guess the next biweekly maintenance announcement will be on 14th december, right?

answered 5 years ago
0

Any update on the subject ?
The problem still occurs and it's start becoming quite problematic for our clients.

answered 5 years ago
0

Any updates on this issue.We have been getting assert error intermittently.Any resolution is much appreciated.

answered 5 years ago
0

Thanks for reporting the issue. The assert error comes from a race condition during the processing of a query and that is why some users are encountering this intermittently. We have a fix for this issue and it is expected to start rolling out in a few weeks. Once a version number for the release containing the fix is confirmed, I will post it here for your awareness.

answered 5 years ago
0

If you are experiencing significant pain due to this issue and waiting (about 10 business days) for an official patch to roll-out impacts your workload, please send me a private message and we can explore the option to private patch your cluster with the fix.

answered 5 years ago
0

The fix for this issue has been deployed on Jan 11, 2018, please see the release announcement here: https://forums.aws.amazon.com/ann.jspa?annID=5358

If your cluster is on a build level that is 1.0.1636 or higher, you should no longer encounter this issue. If would be great if someone can validate their scenario against this build level.

Thank you.

answered 5 years ago
0

Hello,
Our Redshift cluster is still on version 1.0.1582
Our maintenance window is Wed. 8:00-8:30

Is there any way you could patch our cluster this week?

We've been having a lot of errors lately while running our data pipelines, which we hope will get solved by this patch.

error: Assert
code: 1000
context: p - id=334 not in rowsets list
...

Thanks for your help.
-Jeff

answered 5 years ago
0

Hi Jeff,

What region is your cluster in? If you private message me your cluster endpoint, we can check when the patch with the fix will be available for your region.

answered 5 years ago
0

In case anyone is having the same issue and wondering about the timing of patch updates...
The updates are rolled out by region.
I'll update the post with results of whether we're still seeing the assert errors or not once we get the patch.
Thanks for your help Naresh.
Best,
-Jeff

The us-east-1 region will get the patch with a fix
this Thursday by noon Pacific Time. If you change
your maintenance window to Thursday evening, you will
get the fix in a couple of days.

answered 5 years ago
0

Hi Naresh - just looking at this thread as we had an assert error begin to pop up on Friday 1/26, and am wondering if it's related to this (or possibly even triggered by the fix). We've gotten two variations in the error:

"error: Assert
code: 1000
context: mod == -1 || (mod - VARHDRSZ) == size -"

and

"error: Assert
code: 1000
context: val->len <= maxlen -"

I've confirmed that I can run the relevant model in a SQL client but it will not run as part of our scheduled process (ran completley fine until 1/26, with no changes since). Please let me know if you're able to assist!

answered 5 years ago
0

Hi Mike,

Based on a quick look, this new assert seems unrelated to what was originally reported in this thread. Could you please do one of following:

  1. Open a support case with all pertinent information for us to investigate this error
  2. Start a new forum post to discuss this

With either option, if you can provide us with the relevant DDL and the query so we can reproduce the problem in the lab, that would enable us to investigate and try to provide you with a workaround to avoid this issue.

Thanks.

Edited by: awsnaresh on Jan 30, 2018 9:20 AM

answered 5 years ago
0

Hi Naresh,
Yes the patch is definitely working and seems to have fixed all the assert errors we were seeing before. We just ran about 60 pipeline jobs over the weekend and none of them produced the assert error. So it looks like this particular issue is fixed.

error: Assert
code: 1000
context: p - id=334 not in rowsets list

Cluster Version
Version number of Redshift for this cluster. 1.0.1657

Thanks for your help.
-Jeffery

answered 5 years ago
0

Great, thanks for confirming, Jeff!

answered 5 years ago
0

I solved this by materialising the upstream views as tables and then selecting from them.
It's a bit of a bandaid but worth a shot.

Edited by: dominic2 on Oct 31, 2019 7:05 PM

answered 3 years ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions