Tier0 logError skim in Stage2

[rekovic]

The "beam mode" information coming from BST should be available in TCDS event record;
https://twiki.cern.ch/twiki/bin/view/CMS/TcdsEventRecord#TCDS_Data_Block_Definition_curre

Best regards,
Takashi

2016/02/22 11:57、Michael Mulhearn [email protected] :

Vladimir has kindly agreed to look into this.

Takashi may even already know if this available in e.g. External
Conditions block from new uGT...

-Mike

On Mon, Feb 22, 2016 at 11:50 AM, Slava Krutelyov [email protected] wrote:

On 2/22/16 2:09 AM, Michael Mulhearn wrote:

We no longer have gtEvmDigis for sure. Adding them back in will not
help, as there will be nothing in the RAW data to unpack.

Its hard for me to believe this is needed functionality, since we ran
all of Stage-1 without them!

Hi Mike,

logError skim used a stable beam flag to have a cleaner output.
In the past it was derived from the gtEvmDigis.
Do you happen to know what would give this information now?

           --slava

-Mike

On Mon, Feb 22, 2016 at 8:48 AM, David Lange [email protected] wrote:

Hi Slava (+Mike)

On Feb 21, 2016, at 9:28 PM, Slava Krutelyov [email protected] wrote:

On 2/21/16 9:49 AM, David Lange wrote:

Also, a general issue for skims: LogError* skims rely on the stableBeam

filter, which is based on gtEvmDigis data. I reintroduced gtEvmDigis
(see the github link above),
but I have no idea if it will run.
I've sent a message to
https://hypernews.cern.ch/HyperNews/CMS/get/L1TriggerSW/583.html,
but didn't get an answer.
Mike?
you were pinged on this a 6-8 months back:)

I have apparently missed the ping.
This now comes as a reminder then.

(was not a ping from me… but I remember one that I decided not to care about)

• we should remove gtEvmDigis - I don’t have any memory about why a
  log-error skim would want to depend on it - but its gone with stage 1
  trigger…

The reason is that outside of the stable beams period chances are fairly
high that we will have more errors.
So, the logerror skim will be polluted.
Is there an alternative in the currently available data (TCDS or other
pieces)?

Apparently, this also means that the logerror skim wasn't working in 2015.
Looking at 2015{C,D}, I see that files are empty.
I don't really see logerror in 2015B, it was probably not in the configs.

the L1 would have changed after 2015A.. so thats consistent.. but a replacement is indeed a question for Mike(?)

David

         --slava