[jira] [Commented] (TS-4778) CID 1361874: Null pointer dereferences (REVERSE_INULL) /mgmt/api/CfgContextManager.cc:
[ https://issues.apache.org/jira/browse/TS-4778?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15447106#comment-15447106 ] Eric Schwartz commented on TS-4778: --- This is from Coverity? Do we know where it's being dereferenced? > CID 1361874: Null pointer dereferences (REVERSE_INULL) > /mgmt/api/CfgContextManager.cc: > > > Key: TS-4778 > URL: https://issues.apache.org/jira/browse/TS-4778 > Project: Traffic Server > Issue Type: Bug > Components: Management API >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > Not sure why this showed up now, but it did... > {code} > *** CID 1361874: Null pointer dereferences (REVERSE_INULL) > /mgmt/api/CfgContextManager.cc: 201 in CfgContextGet(CfgContext *)() > 195 ats_free(old_text); // need to free memory > 196 return ret; > 197 } > 198 } > 199 delete (rule_list); // free RuleList memory > 200 // TODO: Hmmm, this looks almost like a memory leak, why the strcmp > ?? >CID 1361874: Null pointer dereferences (REVERSE_INULL) >Null-checking "old_text" suggests that it may be null, but it has already > been dereferenced on all paths leading to the check. > 201 if (old_text && strcmp(old_text, "") != 0) { > 202 ats_free(old_text); // need to free memory > 203 } > 204 return TS_ERR_OKAY; > 205 } > 206 > {code} -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-4530) Enable proxy.config.hostdb.host_file.path by default
[ https://issues.apache.org/jira/browse/TS-4530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15437774#comment-15437774 ] Eric Schwartz edited comment on TS-4530 at 8/25/16 10:07 PM: - This is a string config variable. What should we make the default? I know the Yahoo default but don't know what's best for the general case. Maybe something depending on layout? was (Author: es): This is a string config variable. What should we make the default? I know the Yahoo default but don't know what's best for the general case. > Enable proxy.config.hostdb.host_file.path by default > > > Key: TS-4530 > URL: https://issues.apache.org/jira/browse/TS-4530 > Project: Traffic Server > Issue Type: Improvement > Components: Configuration >Reporter: James Peach >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > {{proxy.config.hostdb.host_file.path}} is disabled by default. For 7.0 we > should enable it do that Traffic Server meets operators expectations in the > default config. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4530) Enable proxy.config.hostdb.host_file.path by default
[ https://issues.apache.org/jira/browse/TS-4530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15437774#comment-15437774 ] Eric Schwartz commented on TS-4530: --- This is a string config variable. What should we make the default? I know the Yahoo default but don't know what's best for the general case. > Enable proxy.config.hostdb.host_file.path by default > > > Key: TS-4530 > URL: https://issues.apache.org/jira/browse/TS-4530 > Project: Traffic Server > Issue Type: Improvement > Components: Configuration >Reporter: James Peach >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > {{proxy.config.hostdb.host_file.path}} is disabled by default. For 7.0 we > should enable it do that Traffic Server meets operators expectations in the > default config. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4423) Don't show function / filename / line numbers on "operational logs"
[ https://issues.apache.org/jira/browse/TS-4423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15437731#comment-15437731 ] Eric Schwartz commented on TS-4423: --- PR here: https://github.com/apache/trafficserver/pull/928 > Don't show function / filename / line numbers on "operational logs" > --- > > Key: TS-4423 > URL: https://issues.apache.org/jira/browse/TS-4423 > Project: Traffic Server > Issue Type: Improvement > Components: Logging >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > As of v6.0.0, we've enabled the configuration > {code} > CONFIG proxy.config.diags.show_location INT 1 > {code} > This means we now show file names / numbers on all logging, including > Notes(), Warning()'s etc. A suggestion is that for 7.0.0, we restore some of > the old functionality, such that we only show location on Debug() when this > is enabled. As an additional option feature, we could add an additional enum > value to this (2), which would then mean show location for all logging. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-4423) Don't show function / filename / line numbers on "operational logs"
[ https://issues.apache.org/jira/browse/TS-4423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15435621#comment-15435621 ] Eric Schwartz edited comment on TS-4423 at 8/24/16 8:28 PM: So to make sure I've got this before I change the code: Current: 0 - don't show any line numbers 1 (DEFAULT) - show for all messages Proposed: 0 - don't show any line numbers 1 (DEFAULT) - show for Debug() messages only 2 - show for all messages was (Author: es): So to make sure I've got this before I change the code: 0 - don't show any line numbers 1 (DEFAULT) - show for Debug() messages only 2 - show for all messages > Don't show function / filename / line numbers on "operational logs" > --- > > Key: TS-4423 > URL: https://issues.apache.org/jira/browse/TS-4423 > Project: Traffic Server > Issue Type: Improvement > Components: Logging >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > As of v6.0.0, we've enabled the configuration > {code} > CONFIG proxy.config.diags.show_location INT 1 > {code} > This means we now show file names / numbers on all logging, including > Notes(), Warning()'s etc. A suggestion is that for 7.0.0, we restore some of > the old functionality, such that we only show location on Debug() when this > is enabled. As an additional option feature, we could add an additional enum > value to this (2), which would then mean show location for all logging. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-4423) Don't show function / filename / line numbers on "operational logs"
[ https://issues.apache.org/jira/browse/TS-4423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15435621#comment-15435621 ] Eric Schwartz edited comment on TS-4423 at 8/24/16 8:27 PM: So to make sure I've got this before I change the code: 0 - don't show any line numbers 1 (DEFAULT) - show for Debug() messages only 2 - show for all messages was (Author: es): So to make sure I've got this before I write the code: 0 - don't show any line numbers 1 (DEFAULT) - show for Debug() messages only 2 - show for all messages > Don't show function / filename / line numbers on "operational logs" > --- > > Key: TS-4423 > URL: https://issues.apache.org/jira/browse/TS-4423 > Project: Traffic Server > Issue Type: Improvement > Components: Logging >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > As of v6.0.0, we've enabled the configuration > {code} > CONFIG proxy.config.diags.show_location INT 1 > {code} > This means we now show file names / numbers on all logging, including > Notes(), Warning()'s etc. A suggestion is that for 7.0.0, we restore some of > the old functionality, such that we only show location on Debug() when this > is enabled. As an additional option feature, we could add an additional enum > value to this (2), which would then mean show location for all logging. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4423) Don't show function / filename / line numbers on "operational logs"
[ https://issues.apache.org/jira/browse/TS-4423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15435621#comment-15435621 ] Eric Schwartz commented on TS-4423: --- So to make sure I've got this before I write the code: 0 - don't show any line numbers 1 (DEFAULT) - show for Debug() messages only 2 - show for all messages > Don't show function / filename / line numbers on "operational logs" > --- > > Key: TS-4423 > URL: https://issues.apache.org/jira/browse/TS-4423 > Project: Traffic Server > Issue Type: Improvement > Components: Logging >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > As of v6.0.0, we've enabled the configuration > {code} > CONFIG proxy.config.diags.show_location INT 1 > {code} > This means we now show file names / numbers on all logging, including > Notes(), Warning()'s etc. A suggestion is that for 7.0.0, we restore some of > the old functionality, such that we only show location on Debug() when this > is enabled. As an additional option feature, we could add an additional enum > value to this (2), which would then mean show location for all logging. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (TS-4778) CID 1361874: Null pointer dereferences (REVERSE_INULL) /mgmt/api/CfgContextManager.cc:
[ https://issues.apache.org/jira/browse/TS-4778?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz reassigned TS-4778: - Assignee: Eric Schwartz > CID 1361874: Null pointer dereferences (REVERSE_INULL) > /mgmt/api/CfgContextManager.cc: > > > Key: TS-4778 > URL: https://issues.apache.org/jira/browse/TS-4778 > Project: Traffic Server > Issue Type: Bug > Components: Management API >Reporter: Leif Hedstrom >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > Not sure why this showed up now, but it did... > {code} > *** CID 1361874: Null pointer dereferences (REVERSE_INULL) > /mgmt/api/CfgContextManager.cc: 201 in CfgContextGet(CfgContext *)() > 195 ats_free(old_text); // need to free memory > 196 return ret; > 197 } > 198 } > 199 delete (rule_list); // free RuleList memory > 200 // TODO: Hmmm, this looks almost like a memory leak, why the strcmp > ?? >CID 1361874: Null pointer dereferences (REVERSE_INULL) >Null-checking "old_text" suggests that it may be null, but it has already > been dereferenced on all paths leading to the check. > 201 if (old_text && strcmp(old_text, "") != 0) { > 202 ats_free(old_text); // need to free memory > 203 } > 204 return TS_ERR_OKAY; > 205 } > 206 > {code} -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (TS-4457) Via header always reports http1
[ https://issues.apache.org/jira/browse/TS-4457?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz reassigned TS-4457: - Assignee: Eric Schwartz > Via header always reports http1 > --- > > Key: TS-4457 > URL: https://issues.apache.org/jira/browse/TS-4457 > Project: Traffic Server > Issue Type: Bug >Reporter: Miles Libbey >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > When using http2, the Via header says http1.1. > $ curl -D- -o/dev/null -s "https://docs.trafficserver.apache.org/; | grep via > via:http/1.1 ATS (ApacheTrafficServer/6.2.0 [cRs f ]) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4012) tcp_reused log field doesn't work for http2
[ https://issues.apache.org/jira/browse/TS-4012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15425182#comment-15425182 ] Eric Schwartz commented on TS-4012: --- TS-3612 seems to have fixed this because HttpSM simply has access to the underlying netvc from the parent session. > tcp_reused log field doesn't work for http2 > --- > > Key: TS-4012 > URL: https://issues.apache.org/jira/browse/TS-4012 > Project: Traffic Server > Issue Type: Bug > Components: HTTP/2, Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > This is something we noticed at Yahoo! > Because it uses the HttpClientSession transact_count, the tcp_reused log > field will not properly work for http2 and spdy, where the HttpClientSession > isn't reused. > We need to essentially count the number of initiated streams instead. > We've got a fix that does this for HTTP2. With SPDY eventually dying off, > does it make sense to also submit a change to fix this for SPDY? We have code > to do this as well, but it currently involves an H2 dependency that might be > undesirable. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4012) tcp_reused log field doesn't work for http2
[ https://issues.apache.org/jira/browse/TS-4012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15424915#comment-15424915 ] Eric Schwartz commented on TS-4012: --- [~bcall] running a test today on our internal traffic with TS-3612 to see if that fixed this. (It may have) Will update this with results too. > tcp_reused log field doesn't work for http2 > --- > > Key: TS-4012 > URL: https://issues.apache.org/jira/browse/TS-4012 > Project: Traffic Server > Issue Type: Bug > Components: HTTP/2, Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > This is something we noticed at Yahoo! > Because it uses the HttpClientSession transact_count, the tcp_reused log > field will not properly work for http2 and spdy, where the HttpClientSession > isn't reused. > We need to essentially count the number of initiated streams instead. > We've got a fix that does this for HTTP2. With SPDY eventually dying off, > does it make sense to also submit a change to fix this for SPDY? We have code > to do this as well, but it currently involves an H2 dependency that might be > undesirable. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-4723) ATS CARP Plugin
Eric Schwartz created TS-4723: - Summary: ATS CARP Plugin Key: TS-4723 URL: https://issues.apache.org/jira/browse/TS-4723 Project: Traffic Server Issue Type: New Feature Components: Plugins Reporter: Eric Schwartz Open sourcing this plugin we use internally within Yahoo in place of hierarchical caching. CARP is a plugin that allows you to group a bunch of ATS hosts into a cluster and share cache space across the entire group. This is done with consistent hashing on the object URL to generate an "owner" node in the cluster. Requests to any other node in the cluster will be forwarded on to the corresponding owner. More info in the README. Difference from internal version of note: I've ripped out some code we weren't entirely sure we could open source because of a hash function. If it turns out that we can open source this, I'll do so. The CarpHashAlgorithm class is meant to be extensible, so any consistent hash function can replace it. The function included here is pretty straightforward but not what we use in production, so just wanted to use that caveat. One last caveat: You'll see some code and documentation in here for object replication. This is something I added recently to CARP that allows you to specify an object be replicated a certain number of times in the cluster. This is useful if you have a network partition or if you're performing some sort of update. When an object's primary owner is unreachable, a node in the cluster can go to the secondary owner if it's available rather than having to fall all the way back to origin. While I've done some initial testing on this with my own cluster of hosts, it's not been tested in production so use at your own risk for now. I'll be sure to keep the open source community informed on the progress of our tests with this feature. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (TS-4723) ATS CARP Plugin
[ https://issues.apache.org/jira/browse/TS-4723?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz reassigned TS-4723: - Assignee: Eric Schwartz > ATS CARP Plugin > --- > > Key: TS-4723 > URL: https://issues.apache.org/jira/browse/TS-4723 > Project: Traffic Server > Issue Type: New Feature > Components: Plugins >Reporter: Eric Schwartz >Assignee: Eric Schwartz > > Open sourcing this plugin we use internally within Yahoo in place of > hierarchical caching. > CARP is a plugin that allows you to group a bunch of ATS hosts into a cluster > and share cache space across the entire group. This is done with consistent > hashing on the object URL to generate an "owner" node in the cluster. > Requests to any other node in the cluster will be forwarded on to the > corresponding owner. More info in the README. > Difference from internal version of note: > I've ripped out some code we weren't entirely sure we could open source > because of a hash function. If it turns out that we can open source this, > I'll do so. The CarpHashAlgorithm class is meant to be extensible, so any > consistent hash function can replace it. The function included here is pretty > straightforward but not what we use in production, so just wanted to use that > caveat. > One last caveat: > You'll see some code and documentation in here for object replication. This > is something I added recently to CARP that allows you to specify an object be > replicated a certain number of times in the cluster. This is useful if you > have a network partition or if you're performing some sort of update. When an > object's primary owner is unreachable, a node in the cluster can go to the > secondary owner if it's available rather than having to fall all the way back > to origin. While I've done some initial testing on this with my own cluster > of hosts, it's not been tested in production so use at your own risk for now. > I'll be sure to keep the open source community informed on the progress of > our tests with this feature. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4012) tcp_reused log field doesn't work for http2/spdy
[ https://issues.apache.org/jira/browse/TS-4012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15385052#comment-15385052 ] Eric Schwartz commented on TS-4012: --- Yes! I should be able to fix this now that TS-3612 is finalized. On Tuesday, July 19, 2016 4:12 PM, Leif Hedstrom (JIRA)wrote: [ https://issues.apache.org/jira/browse/TS-4012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15385044#comment-15385044 ] Leif Hedstrom commented on TS-4012: --- [~erics273]Are we getting a new PR for this? Or should we close this? -- This message was sent by Atlassian JIRA (v6.3.4#6332) > tcp_reused log field doesn't work for http2/spdy > > > Key: TS-4012 > URL: https://issues.apache.org/jira/browse/TS-4012 > Project: Traffic Server > Issue Type: Bug > Components: HTTP/2, Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz > Fix For: 7.0.0 > > > This is something we noticed at Yahoo! > Because it uses the HttpClientSession transact_count, the tcp_reused log > field will not properly work for http2 and spdy, where the HttpClientSession > isn't reused. > We need to essentially count the number of initiated streams instead. > We've got a fix that does this for HTTP2. With SPDY eventually dying off, > does it make sense to also submit a change to fix this for SPDY? We have code > to do this as well, but it currently involves an H2 dependency that might be > undesirable. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-4379) Log field for connections to origin
Eric Schwartz created TS-4379: - Summary: Log field for connections to origin Key: TS-4379 URL: https://issues.apache.org/jira/browse/TS-4379 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Eric Schwartz Simple log field for # of connections open to the origin server. Did this a while back for some work we were doing to track the maximum number of connections to a given origin across some traffic spikes. Thought someone else might find it useful. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4070) RemapProcessor Forward Mapping with Recv Port failing with TS-2157 changes
[ https://issues.apache.org/jira/browse/TS-4070?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15060997#comment-15060997 ] Eric Schwartz commented on TS-4070: --- Ah good catch, sorry about the mixup. I had one comment on your PR but other than that looks like a good fix to me too. > RemapProcessor Forward Mapping with Recv Port failing with TS-2157 changes > -- > > Key: TS-4070 > URL: https://issues.apache.org/jira/browse/TS-4070 > Project: Traffic Server > Issue Type: Bug > Components: Network >Reporter: Craig Schomburg >Assignee: Eric Schwartz >Priority: Critical > Labels: regression > Fix For: 6.1.0 > > Attachments: TS-4070.patch > > > During the rework of RemapProcessor.cc, RemapProcessor::setup_for_remap() as > part of the TS-2157 changeset, the port access API appears to have been > incorrectly modified to use the client_info.src_addr.host_order_port() API > [source port, host order] instead of the client_info.dst_addr.port() > [destination/receive port, network > order] API. This caused port based remapping based on the receive port to > fail with ATS 6.0.0. > Functionality was previously working with ATS 4.0.1. > See attached patch that was used to restore the ATS 4.0.1 functionality. > Looking for confirmation that this fix/patch was correct. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-4031) Update docs to show new log fields
Eric Schwartz created TS-4031: - Summary: Update docs to show new log fields Key: TS-4031 URL: https://issues.apache.org/jira/browse/TS-4031 Project: Traffic Server Issue Type: Bug Components: Documentation Reporter: Eric Schwartz Lost some log fields a few weeks ago in an update. Need to get them back. Feel free to assign this to me. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-4024) wire tracing enhancements
Eric Schwartz created TS-4024: - Summary: wire tracing enhancements Key: TS-4024 URL: https://issues.apache.org/jira/browse/TS-4024 Project: Traffic Server Issue Type: Bug Components: Logging Reporter: Eric Schwartz Some enhancements from Yahoo! to make wire tracing dynamically enable-able and to enhance log messages. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4024) wire tracing enhancements
[ https://issues.apache.org/jira/browse/TS-4024?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15004871#comment-15004871 ] Eric Schwartz commented on TS-4024: --- Could someone assign this to me? > wire tracing enhancements > - > > Key: TS-4024 > URL: https://issues.apache.org/jira/browse/TS-4024 > Project: Traffic Server > Issue Type: Bug > Components: Logging >Reporter: Eric Schwartz > > Some enhancements from Yahoo! to make wire tracing dynamically enable-able > and to enhance log messages. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-4024) wire tracing enhancements
[ https://issues.apache.org/jira/browse/TS-4024?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15004871#comment-15004871 ] Eric Schwartz edited comment on TS-4024 at 11/13/15 11:06 PM: -- Could someone assign this to me? Have a patch for discussion. was (Author: es): Could someone assign this to me? > wire tracing enhancements > - > > Key: TS-4024 > URL: https://issues.apache.org/jira/browse/TS-4024 > Project: Traffic Server > Issue Type: Bug > Components: Logging >Reporter: Eric Schwartz > > Some enhancements from Yahoo! to make wire tracing dynamically enable-able > and to enhance log messages. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-4012) tcp_reused log field doesn't work for http2/spdy
[ https://issues.apache.org/jira/browse/TS-4012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14999396#comment-14999396 ] Eric Schwartz commented on TS-4012: --- Also, could someone assign this to me? > tcp_reused log field doesn't work for http2/spdy > > > Key: TS-4012 > URL: https://issues.apache.org/jira/browse/TS-4012 > Project: Traffic Server > Issue Type: Bug > Components: Logging >Reporter: Eric Schwartz > > This is something we noticed at Yahoo! > Because it uses the HttpClientSession transact_count, the tcp_reused log > field will not properly work for http2 and spdy, where the HttpClientSession > isn't reused. > We need to essentially count the number of initiated streams instead. > We've got a fix that does this for HTTP2. With SPDY eventually dying off, > does it make sense to also submit a change to fix this for SPDY? We have code > to do this as well, but it currently involves an H2 dependency that might be > undesirable. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-4012) tcp_reused log field doesn't work for http2/spdy
Eric Schwartz created TS-4012: - Summary: tcp_reused log field doesn't work for http2/spdy Key: TS-4012 URL: https://issues.apache.org/jira/browse/TS-4012 Project: Traffic Server Issue Type: Bug Components: Logging Reporter: Eric Schwartz This is something we noticed at Yahoo! Because it uses the HttpClientSession transact_count, the tcp_reused log field will not properly work for http2 and spdy, where the HttpClientSession isn't reused. We need to essentially count the number of initiated streams instead. We've got a fix that does this for HTTP2. With SPDY eventually dying off, does it make sense to also submit a change to fix this for SPDY? We have code to do this as well, but it currently involves an H2 dependency that might be undesirable. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14933788#comment-14933788 ] Eric Schwartz commented on TS-3900: --- this can be closed out > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz >Priority: Minor > Labels: yahoo > Fix For: 6.1.0 > > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3648) Desire support for client TLS cipher in custom log format
[ https://issues.apache.org/jira/browse/TS-3648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14933789#comment-14933789 ] Eric Schwartz commented on TS-3648: --- this can be closed out > Desire support for client TLS cipher in custom log format > - > > Key: TS-3648 > URL: https://issues.apache.org/jira/browse/TS-3648 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Sproul >Assignee: Eric Schwartz > Fix For: 6.1.0 > > > While the TLS cipher stats from TS-2169 are awesome for seeing what ciphers > are in active use, I would also like to be able to create a custom log to > track cipher by client request. That would enable me, for example, to > determine who among my users might be affected by a cipher list change. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3911) New log tag for proxy connection being over SSL, pqssl
[ https://issues.apache.org/jira/browse/TS-3911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14933786#comment-14933786 ] Eric Schwartz commented on TS-3911: --- this can be closed out > New log tag for proxy connection being over SSL, pqssl > -- > > Key: TS-3911 > URL: https://issues.apache.org/jira/browse/TS-3911 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz > Fix For: 6.1.0 > > > Want to add a log tag identifying whether the connection between the proxy > and origin is over SSL. We have a similar tag for the client connection. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-3911) Create log tag for server connection isSSL
Eric Schwartz created TS-3911: - Summary: Create log tag for server connection isSSL Key: TS-3911 URL: https://issues.apache.org/jira/browse/TS-3911 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Eric Schwartz Want to add a log tag identifying whether the connection between the proxy and origin is over SSL. We have a similar tag for the client connection. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3911) Create log tag for server connection isSSL
[ https://issues.apache.org/jira/browse/TS-3911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14743967#comment-14743967 ] Eric Schwartz commented on TS-3911: --- 1) Could someone assign this to me? 2) PR here: https://github.com/apache/trafficserver/pull/293 > Create log tag for server connection isSSL > -- > > Key: TS-3911 > URL: https://issues.apache.org/jira/browse/TS-3911 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz > > Want to add a log tag identifying whether the connection between the proxy > and origin is over SSL. We have a similar tag for the client connection. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (TS-3911) Create log tag for whether proxy to server connection is ssl
[ https://issues.apache.org/jira/browse/TS-3911?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz updated TS-3911: -- Summary: Create log tag for whether proxy to server connection is ssl (was: Create log tag for server connection isSSL) > Create log tag for whether proxy to server connection is ssl > > > Key: TS-3911 > URL: https://issues.apache.org/jira/browse/TS-3911 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz > > Want to add a log tag identifying whether the connection between the proxy > and origin is over SSL. We have a similar tag for the client connection. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz updated TS-3900: -- Description: There was a log tag % added a while back by [~fpesce] that logged whether the connection between ATS and the client was reused. There's been some interest at Yahoo! in a similar tag for whether the connection between ATS and the origin server was reused. (was: There was a log tag % added a while back by [~fpesce] that logged whether the connection between ATS and the client was reused. There's been some interest at Yahoo! in a similar tag for whether the connection between ATS and the origin server was reused.) > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14738847#comment-14738847 ] Eric Schwartz commented on TS-3900: --- Don't seem to be able to assign this to myself but I can do this. My plan was to use the transact_count in the HttpServerSession and set the tag based on its count in HttpSM::attach_server_session. There's been some discussion between myself, [~amc] and [~dcarlin] about whether the tag should be a boolean (not reused or reused) or simply a count. Might be nice to do a bool just because it would be consistent with % but wonder what everyone thinks. > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz updated TS-3900: -- Labels: yahoo (was: ) > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-3900) Create log tag for server connection reuse
Eric Schwartz created TS-3900: - Summary: Create log tag for server connection reuse Key: TS-3900 URL: https://issues.apache.org/jira/browse/TS-3900 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Eric Schwartz There was a log tag % added a while back by [~fpesce] that logged whether the connection between ATS and the client was reused. There's been some interest at Yahoo! in a similar tag for whether the connection between ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Eric Schwartz updated TS-3900: -- Priority: Minor (was: Major) > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz >Priority: Minor > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14738866#comment-14738866 ] Eric Schwartz commented on TS-3900: --- [~amc] I'll investigate whether the server session is available at the time of logging right now. I know the client session isn't typically. And TS-3539 makes use of the transact_count counter in the client session, not the counter in the server session. That counter is used for % to log whether the client connection is reused. Was proposing to do the same thing for the origin side connection. > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz >Priority: Minor > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3900) Create log tag for server connection reuse
[ https://issues.apache.org/jira/browse/TS-3900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14739248#comment-14739248 ] Eric Schwartz commented on TS-3900: --- Looks like it's possible server_session is gone at log time so we'll need to add another field. > Create log tag for server connection reuse > -- > > Key: TS-3900 > URL: https://issues.apache.org/jira/browse/TS-3900 > Project: Traffic Server > Issue Type: New Feature > Components: Logging >Reporter: Eric Schwartz >Assignee: Eric Schwartz >Priority: Minor > Labels: yahoo > > There was a log tag % added a while back by [~fpesce] that logged > whether the connection between ATS and the client was reused. There's been > some interest at Yahoo! in a similar tag for whether the connection between > ATS and the origin server was reused. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3534) Wiretracing for SSL connections
[ https://issues.apache.org/jira/browse/TS-3534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14711879#comment-14711879 ] Eric Schwartz commented on TS-3534: --- Just looked into this and looks like they aren't in there. Oversight on my part. Will add today. Wiretracing for SSL connections --- Key: TS-3534 URL: https://issues.apache.org/jira/browse/TS-3534 Project: Traffic Server Issue Type: New Feature Components: Logging, Tools Reporter: Eric Schwartz Assignee: Eric Schwartz Fix For: 6.0.0 Opening a ticket for discussion of the wiretracing change I made on our internal version of ATS. The change allows for tracing requests through ATS for: a small percentage of traffic, traffic from a certain IP and/or traffic to a specific origin. These settings can be combined. The main updates are to SSLNetVConnection and UnixNetVConnection (adding the trace logic) and to the Logging APIs (to add the special trace logs). One change is made to HttpSM to allow client and server traces to be associated with one another. [~dcarlin] has some notes from the summit on the initial discussion. I'll add a pull request with actual code for people to look at shortly. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3781) Ability to log the The proxy request server port
[ https://issues.apache.org/jira/browse/TS-3781?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14635479#comment-14635479 ] Eric Schwartz commented on TS-3781: --- Looks like a duplicate of https://issues.apache.org/jira/browse/TS-2152 Ability to log the The proxy request server port Key: TS-3781 URL: https://issues.apache.org/jira/browse/TS-3781 Project: Traffic Server Issue Type: Improvement Components: Logging Reporter: Jonh Wendell Attachments: 0001-TS-3781-Add-the-log-field-pqsp-server-port.patch In addition to field pqsi, it would be good to have the server port available to logging as well. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3648) Desire support for client TLS cipher in custom log format
[ https://issues.apache.org/jira/browse/TS-3648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14603251#comment-14603251 ] Eric Schwartz commented on TS-3648: --- looking at [~acacio]'s PR it looks like there was one change requested. as this is now assigned to me I can make that edits to his PR and open a new PR if he's no longer working on this? or do we want to go a different way with this? i agree with some of the commentary about not adding new fields to HttpSM every time we need to update log fields. there was some discussion of creating a separate logging substruct in HttpSM. i could also take that on if this and a couple of [~fpesce]'s PRs are reviewed and merged and we like that idea. Desire support for client TLS cipher in custom log format - Key: TS-3648 URL: https://issues.apache.org/jira/browse/TS-3648 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Eric Sproul Assignee: Eric Schwartz Fix For: 6.1.0 While the TLS cipher stats from TS-2169 are awesome for seeing what ciphers are in active use, I would also like to be able to create a custom log to track cipher by client request. That would enable me, for example, to determine who among my users might be affected by a cipher list change. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3476) Add a log tag for application protocol
[ https://issues.apache.org/jira/browse/TS-3476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14571426#comment-14571426 ] Eric Schwartz commented on TS-3476: --- As far as I know, there isn't an existing path from the HttpSM back to the ProxyClientSession currently in the code. (If I'm wrong here, please let me know) Was looking into adding one by associating the HttpClientSession (ua_session) w/ its parent ProxyClientSession in FetchSM. However I could only access it toward the end of FetchSM's life cycle, in the final call to process_fetch_read. The path I found was through the PluginVC's read_state. Is there a way to access the underlying HttpClientSession earlier in FetchSM that anyone knows of? Add a log tag for application protocol -- Key: TS-3476 URL: https://issues.apache.org/jira/browse/TS-3476 Project: Traffic Server Issue Type: New Feature Components: Logging, SSL Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 It seems crucial to be able to log which protocol handler was negotiated with ALPN (and perhaps NPN as long as we support it). This could simple be the string that was negotiated by the client/server in the TLS handshake. For example, with HTTP/2, it would be h2 (or h2-14 with some browsers). A suggested log tag name would be %alpn which seems clear enough to me and easy to remember :). -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3476) Add a log tag for APLN/NPN negotiated protocol
[ https://issues.apache.org/jira/browse/TS-3476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14570004#comment-14570004 ] Eric Schwartz commented on TS-3476: --- Looking at the SSLNetVConnection it looks like we can get the ALPN/NPN negotiated protocol from the connection there pretty easily (right now it's stored temporarily and logged to traffic.out. I don't see a reason we couldn't just make this a field in the SSLNetVC class and then grab that from the NetVC in HttpSM for logging. This works out nicely because even though for h2 and SPDY, the HttpSM doesn't have a flag with the protocol version, the SSLNetVC does. That gets us the SSL case. We wouldn't be able to log the protocol from a UnixNetVC in this way though. Is that a problem/something we can come up with another solution for? Alternatively, for both SSL and non-SSL cases, we can get the negotiated protocol on the client from the uppermost ClientSession. This involves introducing some sort of pointer from the underlying HttpClientSession in our SPDY implementation back to its parent session. I've been playing around with doing this for a while though and it gets very hairy, because it's difficult to get a pointer back to the SpdyClientSession from the HttpClientSession in any sort of fashion in time for the HttpSM to grab the protocol from its ua_session in time for logging. In addition to these two approaches, does anyone have any alternative ideas how we might do this? Add a log tag for APLN/NPN negotiated protocol -- Key: TS-3476 URL: https://issues.apache.org/jira/browse/TS-3476 Project: Traffic Server Issue Type: New Feature Components: Logging, SSL Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 It seems crucial to be able to log which protocol handler was negotiated with ALPN (and perhaps NPN as long as we support it). This could simple be the string that was negotiated by the client/server in the TLS handshake. For example, with HTTP/2, it would be h2 (or h2-14 with some browsers). A suggested log tag name would be %alpn which seems clear enough to me and easy to remember :). -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-3476) Add a log tag for APLN/NPN negotiated protocol
[ https://issues.apache.org/jira/browse/TS-3476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14570004#comment-14570004 ] Eric Schwartz edited comment on TS-3476 at 6/2/15 11:49 PM: Looking at the SSLNetVConnection it looks like we can get the ALPN/NPN negotiated protocol from the connection there pretty easily (right now it's stored temporarily and logged to traffic.out). I don't see a reason we couldn't just make this a field in the SSLNetVC class and then grab that from the NetVC in HttpSM for logging. This works out nicely because even though for h2 and SPDY, the HttpSM doesn't have a flag with the protocol version, the SSLNetVC does. That gets us the SSL case. We wouldn't be able to log the protocol from a UnixNetVC in this way though. Is that a problem/something we can come up with another solution for? Alternatively, for both SSL and non-SSL cases, we can get the negotiated protocol on the client from the uppermost ClientSession. This involves introducing some sort of pointer from the underlying HttpClientSession in our SPDY implementation back to its parent session. I've been playing around with doing this for a while though and it gets very hairy, because it's difficult to get a pointer back to the SpdyClientSession from the HttpClientSession in any sort of fashion in time for the HttpSM to grab the protocol from its ua_session in time for logging. In addition to these two approaches, does anyone have any alternative ideas how we might do this? was (Author: es): Looking at the SSLNetVConnection it looks like we can get the ALPN/NPN negotiated protocol from the connection there pretty easily (right now it's stored temporarily and logged to traffic.out. I don't see a reason we couldn't just make this a field in the SSLNetVC class and then grab that from the NetVC in HttpSM for logging. This works out nicely because even though for h2 and SPDY, the HttpSM doesn't have a flag with the protocol version, the SSLNetVC does. That gets us the SSL case. We wouldn't be able to log the protocol from a UnixNetVC in this way though. Is that a problem/something we can come up with another solution for? Alternatively, for both SSL and non-SSL cases, we can get the negotiated protocol on the client from the uppermost ClientSession. This involves introducing some sort of pointer from the underlying HttpClientSession in our SPDY implementation back to its parent session. I've been playing around with doing this for a while though and it gets very hairy, because it's difficult to get a pointer back to the SpdyClientSession from the HttpClientSession in any sort of fashion in time for the HttpSM to grab the protocol from its ua_session in time for logging. In addition to these two approaches, does anyone have any alternative ideas how we might do this? Add a log tag for APLN/NPN negotiated protocol -- Key: TS-3476 URL: https://issues.apache.org/jira/browse/TS-3476 Project: Traffic Server Issue Type: New Feature Components: Logging, SSL Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 It seems crucial to be able to log which protocol handler was negotiated with ALPN (and perhaps NPN as long as we support it). This could simple be the string that was negotiated by the client/server in the TLS handshake. For example, with HTTP/2, it would be h2 (or h2-14 with some browsers). A suggested log tag name would be %alpn which seems clear enough to me and easy to remember :). -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3589) Enhance header_rewrite to support TRANSACT_COUNT as a condition
[ https://issues.apache.org/jira/browse/TS-3589?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14559180#comment-14559180 ] Eric Schwartz commented on TS-3589: --- Does anyone have time to give this a look? Enhance header_rewrite to support TRANSACT_COUNT as a condition --- Key: TS-3589 URL: https://issues.apache.org/jira/browse/TS-3589 Project: Traffic Server Issue Type: New Feature Components: Plugins Reporter: Eric Schwartz Assignee: Eric Schwartz Fix For: 6.0.0 Adding support to header_rewrite to support TRANSACT_COUNT as a condition. Will allow us to add/modify headers on connections that share a Client Session. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3534) Wiretracing for SSL connections
[ https://issues.apache.org/jira/browse/TS-3534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14559188#comment-14559188 ] Eric Schwartz commented on TS-3534: --- Does anyone have time to give this a look? Wiretracing for SSL connections --- Key: TS-3534 URL: https://issues.apache.org/jira/browse/TS-3534 Project: Traffic Server Issue Type: New Feature Components: Logging, Tools Reporter: Eric Schwartz Assignee: Eric Schwartz Fix For: sometime Opening a ticket for discussion of the wiretracing change I made on our internal version of ATS. The change allows for tracing requests through ATS for: a small percentage of traffic, traffic from a certain IP and/or traffic to a specific origin. These settings can be combined. The main updates are to SSLNetVConnection and UnixNetVConnection (adding the trace logic) and to the Logging APIs (to add the special trace logs). One change is made to HttpSM to allow client and server traces to be associated with one another. [~dcarlin] has some notes from the summit on the initial discussion. I'll add a pull request with actual code for people to look at shortly. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3589) Enhance header_rewrite to support TRANSACT_COUNT as a condition
[ https://issues.apache.org/jira/browse/TS-3589?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14538722#comment-14538722 ] Eric Schwartz commented on TS-3589: --- going with TRANSACT-COUNT not TRANSACT_COUNT for consistency* Enhance header_rewrite to support TRANSACT_COUNT as a condition --- Key: TS-3589 URL: https://issues.apache.org/jira/browse/TS-3589 Project: Traffic Server Issue Type: New Feature Components: Plugins Reporter: Eric Schwartz Assignee: Eric Schwartz Fix For: 6.0.0 Adding support to header_rewrite to support TRANSACT_COUNT as a condition. Will allow us to add/modify headers on connections that share a Client Session. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-3589) Enhance header_rewrite to support TRANSACT_COUNT as a condition
Eric Schwartz created TS-3589: - Summary: Enhance header_rewrite to support TRANSACT_COUNT as a condition Key: TS-3589 URL: https://issues.apache.org/jira/browse/TS-3589 Project: Traffic Server Issue Type: New Feature Components: Plugins Reporter: Eric Schwartz Adding support to header_rewrite to support TRANSACT_COUNT as a condition. Will allow us to add/modify headers on connections that share a Client Session. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3534) Wiretracing for SSL connections
[ https://issues.apache.org/jira/browse/TS-3534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14517618#comment-14517618 ] Eric Schwartz commented on TS-3534: --- Opened a PR here for people to review/comment on: https://github.com/apache/trafficserver/pull/194 Wiretracing for SSL connections --- Key: TS-3534 URL: https://issues.apache.org/jira/browse/TS-3534 Project: Traffic Server Issue Type: New Feature Components: Logging, Tools Reporter: Eric Schwartz Assignee: Eric Schwartz Fix For: sometime Opening a ticket for discussion of the wiretracing change I made on our internal version of ATS. The change allows for tracing requests through ATS for: a small percentage of traffic, traffic from a certain IP and/or traffic to a specific origin. These settings can be combined. The main updates are to SSLNetVConnection and UnixNetVConnection (adding the trace logic) and to the Logging APIs (to add the special trace logs). One change is made to HttpSM to allow client and server traces to be associated with one another. [~dcarlin] has some notes from the summit on the initial discussion. I'll add a pull request with actual code for people to look at shortly. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-3534) Wiretracing for SSL connections
Eric Schwartz created TS-3534: - Summary: Wiretracing for SSL connections Key: TS-3534 URL: https://issues.apache.org/jira/browse/TS-3534 Project: Traffic Server Issue Type: New Feature Components: Logging, Tools Reporter: Eric Schwartz Opening a ticket for discussion of the wiretracing change I made on our internal version of ATS. The change allows for tracing requests through ATS for: a small percentage of traffic, traffic from a certain IP and/or traffic to a specific origin. These settings can be combined. The main updates are to SSLNetVConnection and UnixNetVConnection (adding the trace logic) and to the Logging APIs (to add the special trace logs). One change is made to HttpSM to allow client and server traces to be associated with one another. [~dcarlin] has some notes from the summit on the initial discussion. I'll add a pull request with actual code for people to look at shortly. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3396) new_tsqa SPDY protocol selection tests
[ https://issues.apache.org/jira/browse/TS-3396?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14350990#comment-14350990 ] Eric Schwartz commented on TS-3396: --- hey [~jacksontj] how do these changes look to you? new_tsqa SPDY protocol selection tests -- Key: TS-3396 URL: https://issues.apache.org/jira/browse/TS-3396 Project: Traffic Server Issue Type: Test Components: SPDY, tsqa Reporter: Eric Schwartz Assignee: Thomas Jackson Priority: Minor Fix For: sometime Opening this JIRA to submit some SPDY protocol selection tests using spdycat I wrote using the new_tsqa. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3396) new_tsqa SPDY protocol selection tests
[ https://issues.apache.org/jira/browse/TS-3396?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14333631#comment-14333631 ] Eric Schwartz commented on TS-3396: --- Sorry for the confusion w/ 2 PRs here. The first one contained a cert that was unnecessary. I've rewritten the test to simply use one of [~jacksontj]'s certs from test_ssl_multicert, as reflected in the more recent PR. new_tsqa SPDY protocol selection tests -- Key: TS-3396 URL: https://issues.apache.org/jira/browse/TS-3396 Project: Traffic Server Issue Type: Test Components: SPDY, tsqa Reporter: Eric Schwartz Assignee: Thomas Jackson Priority: Minor Fix For: sometime Opening this JIRA to submit some SPDY protocol selection tests using spdycat I wrote using the new_tsqa. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Created] (TS-3396) new_tsqa SPDY protocol selection tests
Eric Schwartz created TS-3396: - Summary: new_tsqa SPDY protocol selection tests Key: TS-3396 URL: https://issues.apache.org/jira/browse/TS-3396 Project: Traffic Server Issue Type: Test Components: SPDY, tsqa Reporter: Eric Schwartz Opening this JIRA to submit some SPDY protocol selection tests using spdycat I wrote using the new_tsqa. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-3396) new_tsqa SPDY protocol selection tests
[ https://issues.apache.org/jira/browse/TS-3396?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14328430#comment-14328430 ] Eric Schwartz commented on TS-3396: --- I have a pr for this. Will submit when not mobile. new_tsqa SPDY protocol selection tests -- Key: TS-3396 URL: https://issues.apache.org/jira/browse/TS-3396 Project: Traffic Server Issue Type: Test Components: SPDY, tsqa Reporter: Eric Schwartz Assignee: Thomas Jackson Priority: Minor Fix For: sometime Opening this JIRA to submit some SPDY protocol selection tests using spdycat I wrote using the new_tsqa. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14251874#comment-14251874 ] Eric Schwartz commented on TS-1461: --- In my testing I never ran into that. How does cop check for manager to be up? The process simply goes into a non signal blocking sleep (nanosleep) for a time, which shouldn't be a problem if cop is just checking if the process is up, right? traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14251882#comment-14251882 ] Eric Schwartz commented on TS-1461: --- Ah ok. Well, in my testing that never seemed to be an issue. I was running all three processes. The timer increased to and then plateaued at 60 seconds I have it logging the times it's waiting for, so this was all visible to me in manager.log. I will admit that I don't know exactly how traffic cop's healthcheck works but can look into it when I'm in to see why I wasn't seeing any issue. traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-2157) Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes
[ https://issues.apache.org/jira/browse/TS-2157?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14250264#comment-14250264 ] Eric Schwartz commented on TS-2157: --- I have started work on this and wanted to bring up the following point/possible ambiguity: In HttpTransact.h we declare 4 sets of ConnectionAttributes. Right now each stores only one addr. For each different set of ConnectionAttributes, it looks like addr corresponds to slightly different things. For example, if we take a connection path to go from the client, to ATS and then on to an origin, client_info's addr is clearly a src_addr while server_info's seems to clearly be a dst_addr. This means the additional field we are adding containing the proxy's info could be either src_addr or dst_addr depending on which ConnectionAttributes struct we are observing. This isn't a problem but is ambiguous (might make additional logging TS-2152 less clear for example). Is this possible ambiguity something we care about? I can totally implement this as src_addr and dst_addr if that's the desired behavior, but would something like proxy_addr and remote_addr make sense instead? So that we're consistently storing the proxy's information in the same IpEndpoint in ConnectionAttributes? Thanks for any input anyone has. Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes - Key: TS-2157 URL: https://issues.apache.org/jira/browse/TS-2157 Project: Traffic Server Issue Type: New Feature Components: Network Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 This would more clearly let us encapsulate the two endpoint's (IpEndpoint) for each connection. In addition, we ought to be able to remove the port member from ConnectionAttributes as well, and its convoluted and overloaded semantics. The appropriate IpEndpoint (src_addr or dst_addr) would hold the port information as necessary. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Work started] (TS-2157) Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes
[ https://issues.apache.org/jira/browse/TS-2157?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Work on TS-2157 started by Eric Schwartz. - Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes - Key: TS-2157 URL: https://issues.apache.org/jira/browse/TS-2157 Project: Traffic Server Issue Type: New Feature Components: Network Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 This would more clearly let us encapsulate the two endpoint's (IpEndpoint) for each connection. In addition, we ought to be able to remove the port member from ConnectionAttributes as well, and its convoluted and overloaded semantics. The appropriate IpEndpoint (src_addr or dst_addr) would hold the port information as necessary. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-2157) Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes
[ https://issues.apache.org/jira/browse/TS-2157?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14250870#comment-14250870 ] Eric Schwartz commented on TS-2157: --- A proposal that I think might be clearer: Instead of src_addr and dst_addr (which gets confusing because of implied directionality) we instead store local_addr and remote_addr. remote_addr would correspond to what addr is now, storing the addr of the remote host to which ats is connecting (both in the case of server_info and client_info to go off of the previous comment's example) local_addr would then store the information for the local IpEndpoint for the ATS host. This would make implementation of TS-2152 (with which i am also conveniently tasked) much more straightforward, I think. Replace addr with appropriate src_addr and dst_addr in ConnectionAttributes - Key: TS-2157 URL: https://issues.apache.org/jira/browse/TS-2157 Project: Traffic Server Issue Type: New Feature Components: Network Reporter: Leif Hedstrom Assignee: Eric Schwartz Fix For: 6.0.0 This would more clearly let us encapsulate the two endpoint's (IpEndpoint) for each connection. In addition, we ought to be able to remove the port member from ConnectionAttributes as well, and its convoluted and overloaded semantics. The appropriate IpEndpoint (src_addr or dst_addr) would hold the port information as necessary. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14240109#comment-14240109 ] Eric Schwartz commented on TS-1461: --- I am working on this now. I think the upper limit on restart time seems reasonable but was wondering if I could get some clarification re: exponential backoff vs # of restarts. I think it definitely makes sense to have it do exponential backoff up to some upper limit but once we reach that limit do we wish to continue attempting to restart or after we get to the point where we've waited 60 seconds for a restart do we want that to be our last attempt? It looks like David's initial concern wouldn't be solved unless we have a limit on the # of times we restart in addition to doing exponential backoff for time between restarts. What do we think would be a reasonable # of attempts? traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14240109#comment-14240109 ] Eric Schwartz edited comment on TS-1461 at 12/9/14 9:43 PM: I am working on this now. I think the upper limit on restart time seems reasonable but was wondering if I could get some clarification re: exponential backoff and # of restarts. I think it definitely makes sense to have it do exponential backoff up to some upper limit but once we reach that limit do we wish to continue attempting to restart or after we get to the point where we've waited 60 seconds for a restart do we want that to be our last attempt? It looks like David's initial concern wouldn't be solved unless we have a limit on the # of times we restart in addition to doing exponential backoff for time between restarts. What do we think would be a reasonable # of attempts? was (Author: es): I am working on this now. I think the upper limit on restart time seems reasonable but was wondering if I could get some clarification re: exponential backoff vs # of restarts. I think it definitely makes sense to have it do exponential backoff up to some upper limit but once we reach that limit do we wish to continue attempting to restart or after we get to the point where we've waited 60 seconds for a restart do we want that to be our last attempt? It looks like David's initial concern wouldn't be solved unless we have a limit on the # of times we restart in addition to doing exponential backoff for time between restarts. What do we think would be a reasonable # of attempts? traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14240246#comment-14240246 ] Eric Schwartz commented on TS-1461: --- Ah, ok. That makes sense to me, but it's definitely different than what David had originally requested w.r.t. to a fixed # of times traffic_manager would attempt the restart. I suppose this would drastically cut down on the number of emails he'd be getting but it wouldn't impose any sort of limit. But just to make sure I understand, you wouldn't want to impose any sort of cap on the # of times it restarted? traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14240246#comment-14240246 ] Eric Schwartz edited comment on TS-1461 at 12/9/14 11:06 PM: - Ah, ok. That makes sense to me, but it's definitely different than what David had originally requested w.r.t. to a fixed # of times traffic_manager would attempt the restart. I suppose this would drastically cut down on the number of emails he'd be getting/stop floods of emails because presumably an admin could respond during one of the waits but it wouldn't impose any sort of limit. But just to make sure I understand, you wouldn't want to impose any sort of cap on the # of times it restarted? was (Author: es): Ah, ok. That makes sense to me, but it's definitely different than what David had originally requested w.r.t. to a fixed # of times traffic_manager would attempt the restart. I suppose this would drastically cut down on the number of emails he'd be getting but it wouldn't impose any sort of limit. But just to make sure I understand, you wouldn't want to impose any sort of cap on the # of times it restarted? traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-1461) traffic_manager should only try X number of times to start traffic_server
[ https://issues.apache.org/jira/browse/TS-1461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14240307#comment-14240307 ] Eric Schwartz commented on TS-1461: --- Spoke to David Carlin on IRC who clarified the discrepancy here. Will implement with exponential back off up to 60 seconds and then just continuing to attempt restarting forever with this maximum wait interval. traffic_manager should only try X number of times to start traffic_server - Key: TS-1461 URL: https://issues.apache.org/jira/browse/TS-1461 Project: Traffic Server Issue Type: Improvement Components: Manager Affects Versions: 3.2.0 Reporter: David Carlin Assignee: Eric Schwartz Priority: Minor Fix For: 5.3.0 If there is a problem with the config and traffic_server doesn't start, traffic_manager still tries indefinitely to restart ATS. This fills up the TS logs, /var/log/messages, and can send out hundreds of emails via proxy.config.alarm_email. Is it possible to have traffic_manager only attempt X number of restarts? Or X number of restarts in a given period of time? -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (TS-2152) Create a tag that logs the destination IP of the client connection
[ https://issues.apache.org/jira/browse/TS-2152?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14207067#comment-14207067 ] Eric Schwartz commented on TS-2152: --- Ping. Create a tag that logs the destination IP of the client connection -- Key: TS-2152 URL: https://issues.apache.org/jira/browse/TS-2152 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Leif Hedstrom Labels: newbie Fix For: 6.0.0 This would log the IP of the proxy server, as the client connected to. This is useful for two reasons: 1) A box can be multi-homed 2) In a IPv4 / IPv6, you can then tell which endpoint the client connected to. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Comment Edited] (TS-2152) Create a tag that logs the destination IP of the client connection
[ https://issues.apache.org/jira/browse/TS-2152?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14207067#comment-14207067 ] Eric Schwartz edited comment on TS-2152 at 11/11/14 9:15 PM: - Commenting so Alan can find my profile. was (Author: es): Ping. Create a tag that logs the destination IP of the client connection -- Key: TS-2152 URL: https://issues.apache.org/jira/browse/TS-2152 Project: Traffic Server Issue Type: New Feature Components: Logging Reporter: Leif Hedstrom Labels: newbie Fix For: 6.0.0 This would log the IP of the proxy server, as the client connected to. This is useful for two reasons: 1) A box can be multi-homed 2) In a IPv4 / IPv6, you can then tell which endpoint the client connected to. -- This message was sent by Atlassian JIRA (v6.3.4#6332)