[squid-users] squid 5.1 missing ENTRY_REQUIRES_COLLAPSING

[Loučanský Lukáš]

Hello,
just rolled out squid 5.1 into production machine and got several of
 
2021/08/11 08:10:45 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr4315=V/0x556f34ceba10*0
    current master transaction: master55
2021/08/11 08:10:48 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr4315=V/0x556f34e7f030*0
    current master transaction: master55
2021/08/11 08:10:49 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr4315=V/0x556f3448d400*0
    current master transaction: master55
2021/08/11 08:10:51 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr4315=V/0x556f375bb810*0
    current master transaction: master55
2021/08/11 08:10:59 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr9536=V/0x556f36dace30*0
    current master transaction: master55

 
kids 1 to 4 are workers. I Was using {process_number} macro to create separate aufs cache_dirs for each worker process, recently have changed to one more rock storage. ie no aufs cache_dirs - only rock with different max-size and slot size. However I'm using {process_number} macro to run store_id_program for each worker - like this
 
#worker1
if ${process_number} = 1
store_id_program /lib/squid5/storeid_file_rewrite /etc/squid5/acl/storeid_rewrite.conf
store_id_children 4 startup=1 idle=1 concurrency=0
acl noStoreID url_regex "/etc/squid5/acl/storeid.whitelist"
store_id_access deny noStoreID
store_id_access allow all
endif

etc. for other workers.
 
 
Seems connected with downloading of MS updates from:
2021/08/11 08:05:04 kid2| ctx: enter level  0: 'http://9.tlu.dl.delivery.mp.microsoft.com/filestreamingservice/files/c11c61e1-2c71-4bdc-be53-3c44d1d491f8?P1=1628662587&P2=404&P3=2&P4=AFBNnEb3I1J1i2txgqn9sgKDx6g4wu4Nk8w9MYCvYZqnQg%2fWduDXekI58E264AhA3%2fpZmM5%2b2%2fPFmER9Vr8xnw%3d%3d'
2021/08/11 08:05:04 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr15985=V/0x556f3450d800*0
    current master transaction: master55
2021/08/11 08:05:38 kid2| ctx: exit level  0
2021/08/11 08:05:38 kid2| ctx: enter level  0: 'http://2.tlu.dl.delivery.mp.microsoft.com/filestreamingservice/files/696b26de-18b8-4bd4-8b77-f7f700102951?P1=1628667955&P2=404&P3=2&P4=B6QCKlBJwUdlz1WE9pHnSFs1fgTTOrjLFl%2fAsdkytWDzyqVeKXqCfdO7IocO2%2bbE1RVMyU7whVyM7iM8MSTs1g%3d%3d'
2021/08/11 08:05:38 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr12742=V/0x556f373e32d0*0
    current master transaction: master55
2021/08/11 08:09:18 kid2| ctx: exit level  0
2021/08/11 08:09:18 kid2| BUG: missing ENTRY_REQUIRES_COLLAPSING for e:tr4315=V/0x556f35068920*0
    current master transaction: master55

storeid_rewrite.conf entry for microsoft.com looks like
^https?:\/\/.+?\.ws\.microsoft\.com\/.+?_([0-9a-z]{40})\.(cab|exe|ms[i|u|f]|asf|wm[v|a]|dat|zip|psf|appx)       http://wupdate.squid.local/$1

Several of 2021/08/10 08:46:54 kid2| assertion failed: Controller.cc:930: "!transients || e.hasTransients()" have been logged yesterday, but none since I swapped to all-rocks.

I'm going to change my config to remove store_id configuration for each worker and instead run it without {process_number} macros. I;m not sure why I did it - I tried to conserver resources - like filedescriptors allocation, dns cache entries etc. not important for diskers - so I created these macro statements to allocate it only for workers.

THX
LL