channel clear protocol warning messages from PCAS
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
EPICS Base |
Fix Released
|
Wishlist
|
Jeff Hill |
Bug Description
I get a lot of this messages into gateway log file (does not matter which version of EPICS base is used).
Do you know what does this mean?
Thank you and best regards
Gasper Jansa
Additional information:
CAS: CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=1 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=2 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=3 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=4 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=5 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=6 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=7 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=8 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=9 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=a CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=b CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=1 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=2 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=3 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=4 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=5 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=6 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=7 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=8 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=9 CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=a CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0 cnt=0 psz=0 avail=b
Feb 16 14:32:04 !!! Errlog message received (message is above) bad resource id in "../../
Feb 16 14:32:04 !!! Errlog message received (message is above)
CAS:
Feb 16 14:32:04 !!! Errlog message received (message is above) bad resource id in "../../
Feb 16 14:32:04 !!! Errlog message received (message is above)
CAS:
Feb 16 14:32:04 !!! Errlog message received (message is above) bad resource id in "../../
Original Mantis Bug: mantis-284
http://
> CAS: CAS Request: epics on cosylab.com: cmd=12 cid=4294967295 typ=0
> cnt=0 psz=0 avail=1
This indicates a client is trying to delete a channel that does not exist in the server. I can think of three scenarios.
In the first one we have a client application that is deleting a channel that has already been deleted. That might even be possible since a channel id is only a pointer and the data structure involved is returned to a free list where it might remain unmolested for some time. If the application deletes the same channel twice the library might be tricked into sending the delete request a second time. The solution will be of course be to fix the client application.
In the second one something is wrong with the client library. If that were the case then the first question to ask is what version of the client library is sending these requests (from epics on cosylab.com). Another possible explanation might be that some client with a protocol other than CA is using the TCP port of the gateway's server.
The third one has the gateway code deleting the PV, and the client library either doesn’t know how to deal with the server deletes the channel protocol, or the client application tries to delete the channel during a short time window existing only while the client library hasn’t yet received the server deletes the channel message.
Since we see "cid=4294967295" on multiple lines one could certainly guess that the client has "issues" involved with repeatedly deleting the same channel. Unfortunately, from this diagnostic we don’t know if this is multiple deletes of the same channel within the same client lifespan or else a client being repeatedly run that somehow always stumbles when attempting to delete the same channel id. That makes it somewhat harder to distinguish between the first two cases above and the third one. Still, I think we can conclude that the timing associated with the 3rd scenario would probably make it less likely to occur repeatedly with the same channel id. So we suspect the first two scenarios, and of those two I would currently place my money on the first one if I had to make a choice, but that guess is mostly based on conjecture.