KAGRA Logbook

DGS (General)

takahiro.yamamoto - 23:51 Saturday 12 July 2025 (34542)

EY TCam server seems to be unreachable

Dear tomorrow's shifters,

According to the beam spot archiver, auto-recovery for TCam viewer isn't working for EY.
EY TCam was alive at least until 11:30 JST.
But it cannot be seen since 11:40 JST.

If it's a problem on Vinagre on k1mon4, it should be recovered automatically.
So it seems to be a problem on PlanetaryImager on cam-eya.
It can be recovered by the procedure of Sec-1.1.5 of DGS manual (JGW-G2516704).
Please try to do it.

-----
We can see that the image layout and resolution has been changed between before 7/11 17:40 JST and after 7/11 17:50 JST.

In usual, camera viewer shows an image with 1/4 resolution with camera specification for reducing a network trafic and CPU load. It's important to avoid a hung up of camera server and/or applications. On the other hand, an image with 1/1 resolution is taken in the TCam session. We know image resolution sometimes is not reverted to 1/4 resolution. (I never reproduced this phenomena by manual request to guardian. So I wonder it comes form the compatibility of TCam guardian code and camera session script.) Anyway, this hang-up seemed to be caused by the large network trafic and heavy CPU load due to un-reverted image resolution after camera session.

Comments to this report:

takahiro.yamamoto - 14:19 Sunday 13 July 2025 (34546)

Hirose-san tried to recover EY TCam, but it didn't come back. So I checked the situation and found that not only the EY TCam server but also the rebooter were unreachable. A server for the laser control of EY Pcal and its rebooter which were connected a same network switch at the EYA booth are reachable. So the network around EYA has no problem. I'm not sure there is a way to recover a trouble on the rebooter itself remotely. If no way, we need to go to EYA for recovering them.

It is not so urgent unless we get into a situation that requires initial alignment. So we can probably wait until weekday. But if we will leave it until Friday, a time table of maintenance tasks will delay roughly 1 hour because TCam photo sessions must wait a recovering work and going to EYA will be probably done after the weekly calibration measurement. Hopefully I want to avoid an extension of whole maintenance time because completing all works by 22:00 is already tight schedule. I guess that the recovering work can be parallelly done with another commissioning activities. So This Tuesday or Thursday seems to be a good day for recovering TCam (though I'm not sure a detailed work plan on this Tuesday and Thursday yet).

satoru.ikeda - 17:56 Sunday 13 July 2025 (34549)

Since the rebooter is connected to the UPS, there is also a suspicion that something might be wrong with the UPS.

shinji.miyoki - 22:05 Sunday 13 July 2025 (34550)

>So This Tuesday or Thursday seems to be a good day for recovering TCam (though I'm not sure a detailed work plan on this Tuesday and Thursday yet).

I agree.